Tesla FSD Without LiDAR: A Dangerous Gamble?

Tesla LiDAR stance accelerates NHTSA investigation into FSD - TheStreet

Consumer Technology Review

March 20, 2026

Special Investigation · Autonomous Vehicles

As federal investigations multiply, jury verdicts pile up, and a fatal crash record grows, the evidence increasingly suggests that Tesla's camera-only approach to autonomous driving is an ideological bet — not an engineering consensus.

March 20, 2026

Bottom Line Up Front: Tesla's Full Self-Driving (FSD) system, which relies exclusively on cameras and neural networks while rejecting LiDAR and radar, has accumulated a damning evidentiary record: nine federally documented crashes linked to degraded camera visibility (including one fatality), three concurrent NHTSA investigations covering 3.2 million vehicles, a landmark $243 million jury verdict for misleading marketing of driver-assist technology, and a robotaxi crash rate estimated at roughly 8× higher than average human drivers — even when safety monitors were present. Every major competitor uses sensor fusion (LiDAR + radar + cameras). Waymo's sensor-rich approach has achieved 6.8× fewer casualty crashes than the human benchmark. Until Tesla can demonstrate, at verifiable scale, that cameras alone match the safety record of multi-sensor systems, the weight of evidence supports treating FSD as a supervised driver-assistance tool — not autonomous driving — and treating the company's camera-only philosophy as an unproven bet with public-safety consequences.

The Philosophical Divide: Cameras vs. Sensor Fusion

No question divides the autonomous-vehicle industry more starkly than this one: do self-driving cars need LiDAR? On one side stands virtually the entire field — Waymo, Zoox, Mercedes-Benz, Toyota, Aurora, and others — all of which deploy multi-sensor architectures combining cameras, radar, and light detection and ranging (LiDAR) arrays. On the other side stands Elon Musk, who has called LiDAR a "fool's errand," dismissing it as "expensive hardware that's worthless on the car."

Tesla's position is not merely a cost-cutting measure — it is a fully articulated engineering philosophy. Musk's argument: humans navigate the world with two eyes; cameras mimic that biological system; therefore cameras, supplemented by powerful AI, should be sufficient. He has extended this logic by pointing to Tesla's enormous fleet data advantage — by early 2026, FSD had accumulated 3.6 billion cumulative miles of driving data, roughly triple what it had logged just a year prior.

But as of March 2026, the real-world record tells a more complicated and troubling story. Federal regulators have escalated their scrutiny of FSD to its highest level yet. Courts have rendered the first major liability verdict against Tesla's driver-assistance technology. And the company's own robotaxi pilot has produced a crash rate that independent analysts say should alarm policymakers and consumers alike.

"The gap between Tesla's autonomous driving claims and the regulatory reality has never been wider." — Electrek, March 19, 2026

What LiDAR Does — and Why It Matters

LiDAR (Light Detection and Ranging) uses pulsed laser light to measure distances and construct detailed, real-time three-dimensional maps of a vehicle's surroundings. Unlike cameras, which interpret depth through image processing and learned inference, LiDAR directly measures distance to objects with centimeter-level precision. Unlike radar, it can resolve fine detail — distinguishing a child from a trash can, or a stopped vehicle from a shadow.

Critically, LiDAR operates independently of ambient light and can function in conditions — fog, bright glare, airborne dust — where camera-based perception degrades significantly. In sensor-fusion architectures, LiDAR data is cross-referenced against camera imagery, a process that allows the system to precisely locate objects in three-dimensional space even when one sensor is compromised.

Sensor Comparison: Key Capabilities

Capability	Camera (Tesla Vision)	Radar	LiDAR
3D Distance Measurement	Inferred (AI)	Limited resolution	Direct / Precise
Performance in Fog/Rain	Severely degraded	Good	Moderate (rain) / Good (fog)
Performance in Bright Glare	Severely degraded	Unaffected	Largely unaffected
Fine Object Detail	Excellent (clear conditions)	Poor	Excellent
Redundancy to Camera	None (same modality)	Yes	Yes (different physics)
Cost (approx. vehicle impact)	Very low	Low	Moderate–high

Waymo's fifth-generation vehicles deploy five LiDARs, six radars, and 29 cameras. When Waymo VP Srikanth Thirumalai presented at the AI4 conference in August 2025, he showed video of LiDAR sensors detecting pedestrians readying to step into a roadway — in both cases before the vehicle's cameras had registered any threat. Both times, the vehicle stopped or maneuvered safely. Thirumalai declined to say directly whether he considered camera-only systems safe for public roads, but said that "objective measures" require safety comparisons at scale, and that claims of sensor-parity needed to be demonstrated, not asserted.

The Federal Investigation: Three Concurrent Probes

As of March 2026, the National Highway Traffic Safety Administration (NHTSA) is running three simultaneous federal investigations into Tesla FSD — an unprecedented level of regulatory scrutiny for a single driver-assistance product.

Investigation 1: Degraded Visibility Crashes (EA26002)

The most significant probe was escalated on March 19, 2026, from a Preliminary Evaluation to an Engineering Analysis — the final investigative step before NHTSA can demand a recall. The investigation, covering an estimated 3.2 million Tesla vehicles equipped with FSD, centers on nine documented crashes in which FSD's degradation detection system allegedly failed to warn drivers when cameras were impaired by sun glare, fog, airborne dust, or other common environmental conditions.

Of the nine incidents, one was fatal. In multiple crashes, NHTSA found that FSD lost track of or failed to detect a lead vehicle entirely. The agency further noted that Tesla had disclosed "internal data and labeling limitations" that may have resulted in under-reporting of similar incidents — meaning the actual number of affected crashes could be higher than currently known. NHTSA also found that a software update Tesla deployed to address the problem may have remediated only three of the nine documented crashes, leaving the remainder unaddressed.

The Engineering Analysis phase — historically the final step before a recall demand — gives Tesla a deadline to submit detailed technical documentation on how FSD's neural networks process degraded visual inputs, what fallback behaviors the system employs when sensor confidence drops, and crash reconstruction data from the incidents under review.

Investigation 2: Traffic Safety Violations (PE25012)

A separate NHTSA preliminary evaluation opened in October 2025 is examining 58 incidents in which FSD vehicles executed maneuvers constituting traffic safety law violations — including proceeding through red traffic signals and driving against the direction of travel on public roadways. The investigation covers approximately 2.9 million vehicles. In six incidents, FSD-engaged vehicles ran red lights and were involved in crashes; four of those crashes resulted in injuries. Multiple incidents of wrong-way driving and illegal lane changes have also been documented.

Investigation 3: Crash Reporting Failures

A third concurrent NHTSA inquiry is examining Tesla's compliance with the agency's Standing General Order requiring crash reporting. Investigators have found that Tesla submitted required crash reports months late in multiple cases. The probe raises questions about whether Tesla's crash data — which the company cites to support FSD safety claims — is complete and timely.

NHTSA Investigation Summary (As of March 2026)

• EA26002 — Degraded visibility crashes; 9 incidents, 1 fatality; covers ~3.2M vehicles; Engineering Analysis phase (pre-recall)
• PE25012 — Traffic violations; 58 incidents, 6 red-light crashes with injuries; covers ~2.9M vehicles; Preliminary Evaluation
• Crash Reporting Probe — Late SGO submissions; ongoing
• Total NHTSA Tesla investigations since 2016: More than 40

The Legal Record: Courts Begin to Hold Tesla Accountable

For years, Tesla successfully deflected litigation over Autopilot and FSD crashes, arguing that the driver — not the system — bore ultimate responsibility, and that the "Supervised" label absolved the company of liability for driver over-reliance. That legal posture suffered major defeats in 2025 and early 2026.

Benavides v. Tesla: The Landmark $243 Million Verdict

In August 2025, a Miami federal jury in Benavides v. Tesla awarded more than $240 million in damages — including $200 million in punitive damages — to the family of 22-year-old Naibel Benavides Leon, killed in a 2019 crash when a Tesla Model S on Autopilot ran a stop sign at approximately 62 miles per hour. The jury found Tesla 33% liable, despite the driver having admitted to distraction, placing 67% of fault on the driver.

The verdict was the first successful trial judgment against Tesla's Autopilot system in the United States. The jury accepted plaintiffs' arguments that Tesla had designed Autopilot to be activatable on unsafe roads, had failed to adequately monitor driver attentiveness, and had engaged in misleading marketing. Critically, recovered vehicle data demonstrated that the Autopilot system had detected obstacles prior to the crash but failed to brake — directly contradicting Tesla's claim that driver behavior was the sole cause.

In February 2026, U.S. District Judge Beth Bloom upheld the verdict, rejecting Tesla's motion for a new trial and ruling that evidence at trial "more than supported" the jury's finding. Tesla has appealed but exhausted its post-trial motions at the district court level.

California Courts and the Branding Question

In December 2025, a California judge ruled that Tesla's use of "Autopilot" in its marketing was misleading and violated state law, calling "Full Self-Driving" a product name that was "actually, unambiguously false and counterfactual." Tesla subsequently discontinued Autopilot as a standalone product in the United States and Canada — a significant admission of the marketing problem the company had long denied.

The Cybertruck FSD Lawsuit and the LiDAR Allegation

In March 2026, a new Texas lawsuit filed in Harris County District Court over a Cybertruck FSD crash on a Houston freeway included an explicit product liability claim over the "absence of LiDAR" as a design defect. The complaint alleged the vehicle was "defective and unreasonably dangerous" due to the lack of LiDAR, an ineffective automatic emergency braking system, inadequate driver monitoring, and misleading marketing. The lawsuit also alleges Tesla negligently hired and retained Elon Musk as CEO — an unusual but legally significant framing that directly attributes the camera-only decision to executive misconduct. The case is pending.

• Apr 2019 Key Largo Fatal Crash Tesla Model S on Autopilot runs stop sign at 62 mph; Naibel Benavides Leon killed. Becomes foundation of landmark liability suit.
• 2021 Tesla Removes Radar Tesla discontinues radar on new vehicles, transitioning to camera-only "Tesla Vision." Driver reports of worsened performance in fog and rain follow.
• Oct 2024 NHTSA Opens PE24031 Preliminary evaluation opened after four FSD crashes in reduced-visibility conditions including one pedestrian fatality.
• Aug 2025 $243M Florida Verdict Miami jury finds Tesla 33% liable in Benavides case; first successful Autopilot trial verdict in U.S. history. Punitive damages: $200M.
• Oct 2025 NHTSA Opens PE25012 New probe opened into FSD traffic safety violations: 58 incidents, including red-light running and wrong-way driving; covers ~2.9M vehicles.
• Dec 2025 California Court: FSD Name "Unambiguously False" California judge rules Tesla's "Full Self-Driving" marketing violates state law. Tesla discontinues Autopilot as standalone product in U.S. and Canada.
• Feb 2026 Florida Verdict Upheld Judge Bloom upholds $243M jury verdict; Tesla's post-trial motions exhausted at district court level. Appeal filed.
• Mar 2026 NHTSA Escalates to Engineering Analysis (EA26002) Probe upgraded to pre-recall phase; covers 3.2M vehicles; 9 crashes documented. Third concurrent FSD investigation. Texas LiDAR defect lawsuit filed.

The Robotaxi Reality Check

Tesla launched its paid robotaxi service in Austin, Texas in June 2025 — initially with human safety monitors in the front passenger seat. The company then began gradually removing monitors from a small number of vehicles starting in January 2026. Elon Musk described the development as Tesla joining "the exclusive club of companies operating truly driverless public transit."

The data behind that milestone, however, was more sobering. By mid-October 2025, Tesla had reported seven crash incidents to NHTSA from its Austin fleet — despite safety monitors whose explicit purpose was preventing additional incidents. Based on Tesla's disclosure that the fleet had traveled approximately 250,000 miles through early November, independent analysts calculated a crash rate of roughly once every 60,000 miles — compared to the average human driver's roughly 500,000 miles between crashes. That implied Tesla's supervised robotaxi fleet was crashing at more than eight times the human rate.

Philip Koopman, an emeritus professor at Carnegie Mellon University and a leading autonomous-systems safety researcher, noted that with a fleet of 30 or fewer vehicles with trained safety supervisors, the number of reportable accidents "should have been fewer than seven." He also observed that Tesla had withheld the narrative descriptions of each crash from NHTSA reports, making independent safety analysis impossible.

Further complicating Tesla's narrative: a reverse-engineering analysis by a Texas A&M engineering student found that the Austin robotaxi service was unavailable roughly 60% of the time, and that only 1–5 vehicles were in active operation simultaneously at most hours — far below Musk's stated goal of 500 vehicles by year-end 2025. Reports also emerged that Tesla's "unsupervised" robotaxis were being followed by trailing Tesla vehicles carrying monitors — suggesting the removal of in-vehicle safety monitors may have been an optics exercise rather than a genuine operational transition.

By contrast, Waymo — operating with LiDAR, radar, and cameras — has accumulated more than 100 million fully driverless miles, achieved 6.8× fewer casualty crashes per million miles than the human benchmark, and operates profitable robotaxi services in Phoenix, San Francisco, Austin, and Los Angeles.

"Tesla's Robotaxi fleet…are crashing more than 8 times as often as human drivers. And that's with a trained safety supervisor in the car ready to intervene." — Electrek, January 22, 2026, citing NHTSA SGO reports and Tesla mileage disclosures

The Expert Consensus — and the Dissent

The autonomous-driving research community is not uniformly opposed to camera-based approaches. Several researchers have noted that the camera-only philosophy could, in theory, be vindicated by sufficiently powerful AI and sufficient training data. Rich Sutton's influential "Bitter Lesson" framework — which argues that general learning systems eventually outperform expert-designed solutions — is frequently invoked by Tesla partisans. And Xpeng Motors in China has also moved toward a camera-first architecture, suggesting the approach is not unique to one eccentric billionaire.

Moreover, Tesla's own safety data, when presented on Tesla's terms, looks impressive: the company claims FSD users travel approximately 2.9 million miles between major collisions, compared to a national average of 505,000 miles. The company's overall crash rate with Autopilot engaged on highways is approximately 8× better than human drivers — a real achievement that deserves acknowledgment.

But these statistics carry important caveats. Highway driving — where Autopilot excels — has a substantially lower baseline crash rate than urban driving, making direct comparisons to the national average misleading. Tesla's three concurrent NHTSA investigations directly question the completeness of the company's crash reporting. And Waymo co-CEO Tekedra Mawakana noted in November 2025 that there are no standardized, comparable safety metrics between the two systems, and challenged the industry to release transparent data: "If you are not being transparent, then…you are not doing what is necessary to earn the right to make the road safer."

System Safety Assessment: Camera-Only vs. Sensor Fusion (Current State, March 2026)

Criterion	Tesla FSD (Camera-Only)	Waymo (Sensor Fusion)
Performance in clear conditions	Strong neural-network object recognition	Strong across all sensor modalities
Performance in degraded visibility	Federal investigation ongoing; 9 documented crashes	LiDAR/radar provide independent visibility channels
Regulatory standing	3 concurrent NHTSA investigations; possible recall	Fully driverless permits in 4+ U.S. cities
Robotaxi safety record	~1 crash/60K mi (supervised); 8× human average	0.41 casualty crashes/M mi; 6.8× better than humans
Scale of driverless operation	~32 vehicles in Austin; ~1–5 active at any time	1,500+ vehicles; 100M+ driverless miles
Transparency / data disclosure	Crash narratives withheld; late SGO reporting investigated	Published detailed annual safety reports; peer-reviewed papers

What Tesla Owners Should Know Right Now

The escalation of NHTSA's visibility-crash investigation to Engineering Analysis status means that regulators now believe sufficient evidence of a safety defect may exist to warrant a recall. While the process could still result in a mandatory over-the-air software update rather than a physical recall, the agency's position is clear: FSD's degradation detection system, in its current and previously updated forms, may fail under common driving conditions.

Consumer Guidance

Owners using FSD (Supervised) should manually disengage the system any time visibility is compromised — including sun glare, fog, dust, heavy rain, or smoke. Do not rely on FSD's degradation alert to warn you in time; NHTSA's investigation has documented nine cases in which that warning arrived too late or not at all. Treat FSD as a Level 2 driver-assistance system requiring constant, active supervision — not as autonomous driving. The term "Full Self-Driving" has been ruled misleading by a California court. Do not allow the name to shape your expectation of the system's capabilities.

The Broader Industry Implication

For the broader autonomous vehicle industry, NHTSA's escalation carries implications beyond Tesla. Every AV developer must ultimately demonstrate, at scale, that their chosen sensor architecture can safely navigate the full range of conditions public roads present. Waymo, Zoox, and Aurora have chosen sensor fusion precisely because it provides independent physics-based redundancy when any single sensing modality fails — the automotive engineering equivalent of defense-in-depth.

Tesla's camera-only architecture is not inherently doomed. Advances in AI, compute, and neural network architecture may eventually close the gap. But as of March 2026, the record indicates that gap is real, measurable, and consequential. An engineering philosophy that was once a bold bet has become the subject of three federal investigations, a landmark jury verdict, a judicial finding of false advertising, and a casualty record that independent researchers have flagged as statistically alarming.

Elon Musk has been right about many things that experts dismissed. He may yet be proven right about cameras. But "may yet" is not a safety standard, and the current evidentiary record does not support the conclusion that Tesla FSD has achieved, or is imminently approaching, the safety profile that autonomous operation of motor vehicles on public roads requires.

Verified Sources & Formal Citations

1. Engineering Analysis EA26002 — Tesla FSD Degraded Visibility Investigation
   National Highway Traffic Safety Administration (NHTSA), Office of Defects Investigation · March 19, 2026
   https://static.nhtsa.gov/odi/inv/2024/INIM-PE24031-62887.pdf
2. Preliminary Evaluation PE25012 — Tesla FSD Traffic Safety Violations
   NHTSA Office of Defects Investigation · October 7, 2025
   https://static.nhtsa.gov/odi/inv/2025/INOA-PE25012-19171.pdf
3. NHTSA is one step away from having to recall FSD in visibility crash probe
   Electrek · March 19, 2026
   https://electrek.co/2026/03/19/nhtsa-upgrades-tesla-fsd-visibility-investigation-3-2-million-vehicles/
4. NHTSA Escalates Tesla FSD Investigation After Additional Crashes
   CBT News · March 19, 2026
   https://www.cbtnews.com/nhtsa-escalates-tesla-fsd-investigation/
5. NHTSA Upgrades Tesla FSD Probe One Step Short of Recall
   Automotive World · March 19, 2026
   https://www.automotiveworld.com/news/nhtsa-upgrades-tesla-fsd-probe-one-step-short-of-recall/
6. Tesla LiDAR Stance Accelerates NHTSA Investigation into FSD
   TheStreet · March 20, 2026
   https://www.thestreet.com/automotive/tesla-lidar-stance-accelerates-nhtsa-investigation-into-fsd
7. Tesla Has to Pay Historic $243 Million Judgement Over Autopilot Crash, Judge Says
   Electrek · February 20, 2026
   https://electrek.co/2026/02/20/tesla-has-to-pay-historical-243-million-judgement-over-autopilot-crash-judge-says/
8. Jury Orders Tesla to Pay More Than $240 Million in Autopilot Crash
   NPR · August 2, 2025
   https://www.npr.org/2025/08/02/nx-s1-5490930/tesla-autopilot-crash-jury-240-million-florida
9. Benavides v. Tesla: A Defense-Side Perspective on Florida's Landmark Autopilot Verdict
   Walsworth LLP (WSHB Law) · 2025
   https://www.wshblaw.com/publication-benavides-v-tesla-a-defense-side-perspective-on-floridas-landmark-autopilot-verdict
10. Tesla Cybertruck Owner Sues Over FSD Crash, Alleges 'Negligent' Retention of Musk
    Electrek · March 11, 2026
    https://electrek.co/2026/03/11/tesla-cybertruck-fsd-lawsuit-musk-negligent-hiring/
11. Tesla Starts Robotaxi Rides Without Safety Monitor in Austin
    Electrek · January 22, 2026
    https://electrek.co/2026/01/22/tesla-starts-robotaxi-rides-without-safety-monitor-in-austin-what-you-need-to-know/
12. Tesla's Robotaxi Project in Austin Is Much Smaller Than Musk Claims
    Electrek · December 22, 2025
    https://electrek.co/2025/12/22/tesla-robotaxi-project-austin-much-smaller-than-musk-claims/
13. Tesla Stocks Driverless Robotaxi Tests in Austin — Philip Koopman Comments
    CNBC · December 15, 2025
    https://www.cnbc.com/2025/12/15/tesla-tests-driverless-cars-in-austin-without-humans-on-board.html
14. Tesla's Robotaxi Launches in Austin with Safety Drivers in Passenger Seat
    KVUE (ABC Austin) · June 2025
    https://www.kvue.com/article/money/cars/austin-tesla-robotaxi-launch/269-9d0118a0-a22a-486e-ac6c-a23b84e45d33
15. Waymo Experimenting with Generative AI, but Exec Says LiDAR and Radar Sensors Important to Self-Driving Safety 'Under All Conditions'
    Fortune · August 15, 2025
    https://fortune.com/2025/08/15/waymo-srikanth-thirumalai-interview-ai4-conference-las-vegas-lidar-radar-self-driving-safety-tesla/
16. Tesla Releases Detailed Safety Report After Waymo Co-CEO Called for More Data
    TechCrunch · November 14, 2025
    https://techcrunch.com/2025/11/14/tesla-releases-detailed-safety-report-after-waymo-co-ceo-called-for-more-data/
17. Waymo and Tesla's Self-Driving Systems Are More Similar Than People Think
    Understanding AI (Timothy B. Lee) · December 17, 2025
    https://www.understandingai.org/p/waymo-and-teslas-self-driving-systems
18. Camera versus LiDAR: Waymo vs. Tesla Compared
    The Last Driver License Holder · July 1, 2025
    https://thelastdriverlicenseholder.com/2025/06/25/camera-versus-lidar/
19. Tesla's Big Bet: Cameras Over LiDAR for Self Driving Cars
    Vik's Newsletter · November 17, 2024
    https://www.viksnewsletter.com/p/teslas-big-bet-cameras-over-lidar
20. Tesla Bet on 'Pure Vision' for Self-Driving. That's Why It's in Hot Water
    InsideEVs · October 22, 2024
    https://insideevs.com/news/738204/tesla-pure-vision-camera-only/
21. Tesla Autopilot Hardware — Wikipedia
    Wikipedia, citing primary Tesla and FCC filings · Accessed March 2026
    https://en.wikipedia.org/wiki/Tesla_Autopilot_hardware
22. NHTSA Launches New Tesla 'Full Self-Driving' Investigation on Nearly 2.9 Million Vehicles
    Repairer Driven News · October 10, 2025
    https://www.repairerdrivennews.com/2025/10/10/nhtsa-launches-new-tesla-full-self-driving-investigation-on-nearly-2-9-million-vehicles/
23. Tesla's Self-Driving Ambitions Hit a Wall: NHTSA Probe Puts a March 2026 Deadline on Answers
    WebProNews · March 2026
    https://www.webpronews.com/teslas-self-driving-ambitions-hit-a-wall-nhtsa-probe-puts-a-march-2026-deadline-on-answers/
24. Tesla vs. Waymo — Who Is Closer to Level 5 Autonomous Driving?
    Think Autonomous · September 10, 2025
    https://www.thinkautonomous.ai/blog/tesla-vs-waymo-two-opposite-visions/
25. Mark Rober Tesla Autopilot vs. LiDAR Comparison Video — Analysis
    Electrek · March 23, 2025
    https://electrek.co/2025/03/23/everyones-missing-the-point-of-the-tesla-vision-vs-lidar-wile-e-coyote-video/

This article was prepared for the IPCSG Consumer Technology Review using publicly available federal regulatory filings, court records, peer-reviewed research, and reporting from established technology and automotive publications. All claims have been cross-referenced against primary sources where available. The author has no financial relationship with any automotive or sensor manufacturer. Tesla did not respond to requests for comment on the NHTSA investigation as of the date of publication.

© 2026 Consumer Technology Review — Independent, Unsponsored, Fact-Checked

 

Ballast in the Nose: The Pentagon Accepts Combat-Incapable F-35s as Radar Program Collapses Behind Schedule

EXCLUSIVE: US poised to accept new F-35s without radars, sources say - Breaking Defense

A systemic failure of concurrent development planning has left the world's most expensive weapons program delivering blind fighters to U.S. forces. The history books called this one before.

■ BLUF — Bottom Line Up Front

Beginning as early as June 2025, and accelerating through fall 2026, the U.S. military is accepting production F-35s with weighted ballast in place of an operational radar. The aircraft cannot be combat-coded. The cause is a textbook concurrent-development failure: new Lot 17–19 airframes were structurally redesigned for the next-generation AN/APG-85 radar before that radar was ready — making retrograde installation of the existing APG-81 physically impossible. With Block 4 modernization at least five years behind schedule, acquisition costs exceeding $485 billion, and a GAO finding that contractors collected hundreds of millions in incentive fees while delivering every aircraft in 2024 an average of 238 days late, the APG-85 delay is not an isolated technical glitch — it is the logical endpoint of a program that has chronically overpromised and underdelivered since 2001.

March 19, 2026  —  Updated 17:00 ET

Starting this fall, every new F-35 Lightning II delivered to the United States military will carry a weighted ballast in the nose where its fire-control radar should be. The aircraft will be airworthy. They will not be combat-ready. And depending on how long delays continue, more than 100 jets could enter the inventory in this state — training assets masquerading as frontline fighters at a moment when the Pentagon faces its most demanding strategic environment since the Cold War.

The proximate cause is a structural incompatibility between the new AN/APG-85 radar, developed by Northrop Grumman as the cornerstone of the F-35's Block 4 modernization, and the redesigned forward fuselage bulkhead introduced in Lot 17 production. The two systems were engineered in parallel. The airframe won the race. The radar did not.

As Breaking Defense first exclusively reported on March 19, 2026, the Marine Corps is expected to become the first service to accept F-35Bs without radar, followed later in 2026 by the Air Force and Navy with their F-35A and F-35C variants, respectively. Rep. Rob Wittman (R-Va.), chairman of the House Armed Services Committee's Tactical Air and Land Forces Subcommittee, confirmed the situation publicly, telling reporters: "Right now, they're going to be produced with ballasts, which creates an aircraft that's not going to be combat-coded anytime soon."

■ Program Data At A Glance — F-35 / APG-85

Total Program Lifecycle Cost> $2.0 Trillion (77-year estimate)

Acquisition Cost to Date$485+ Billion (Dec. 2023)

Block 4 Cost Growth$10.6B → $16.5B (+$6B gap, 2021 est.)

Block 4 Schedule Slip≥ 5 years (completion ≥ 2031)

Avg. Delivery Lateness (2024)238 days (all 110 aircraft late)

APG-85 First Intended LotLot 17 (deliveries began 2025)

APG-85 Revised Earliest FieldingLot 20 (deliveries begin 2028)

Contractor Incentive Fees PaidHundreds of millions — despite late delivery

The Engineering Root Cause

To understand the current impasse, one must understand the forward fuselage bulkhead — the structural component that positions and aligns the radar array within the nose of the aircraft. The alignment matters enormously: it establishes the physical attitude of the antenna face, which determines the precision and geometry of the radar beam. Wittman put it plainly: "The bulkhead configuration allows the placement of the radar towards the attitude of the array, and the attitude of the array makes all the difference in the world about how the radar operates."

The APG-81, which has equipped all operational F-35s since the program's early production lots, uses one bulkhead geometry. The APG-85 uses a different one. When the program committed to introducing the new radar beginning with Lot 17, the airframe was physically redesigned to accept the new sensor. That committed the production line to a configuration that cannot accept the older radar as an interim solution — there is no plug-and-play fallback. As one industry source told Defense Daily, aircraft delivered since June 2025 have "APG-85 mountings, which do not fit the APG-81."

The result: jets rolling off the Lockheed Martin production line in Fort Worth with a nose ballast weight to replicate the center-of-gravity contribution of the missing radar. The aircraft can fly. They can even operate in a limited sense alongside radar-equipped jets, receiving targeting data via the F-35's Multifunction Advanced Data Link (MADL) and Link 16 datalinks. But they cannot function independently as combat aircraft. Stacie Pettyjohn, director of the defense program at the Center for a New American Security, characterized an aircraft without a radar as "very near-sighted," noting that while off-board datalinks provide some situational awareness, doing so "would come with some latency and the risk of an enemy jamming communications channels."

"Lots of aircraft out there, but not ones that are ready to go to the fight."

— Rep. Rob Wittman (R-Va.), Chairman, HASC Tactical Air & Land Forces Subcommittee, March 2026

What the APG-85 Was Supposed to Deliver

The stakes of this delay cannot be separated from an understanding of what the APG-85 is meant to accomplish. The new sensor represents a generational transition in active electronically scanned array technology, moving from the gallium arsenide (GaAs) transmit/receive modules of the APG-81 — whose design heritage traces to the 1990s — to a gallium nitride (GaN) architecture that delivers substantially higher power density, improved thermal efficiency, and extended detection range against low-observable targets.

The APG-85 is expected to demand approximately 82 kilowatts of power — a figure that itself drives the separate and also-delayed Engine Core Upgrade (ECU) for the Pratt & Whitney F135 engine. It is designed to provide long-range detection and tracking of aircraft, missiles, and unmanned systems; high-resolution synthetic aperture radar (SAR) ground mapping for targeting and bomb damage assessment; enhanced electronic countermeasures resistance through low probability of intercept/low probability of detection (LPI/LPD) waveform design; and offensive electronic attack capability specifically oriented toward countering integrated air defense networks of the type fielded by China and Russia.

Pettyjohn underscored the strategic dimension: "F-35s have performed really well in the Middle East and in Venezuela against lesser adversaries. But when it comes to countries like China that have built truly integrated air and missile defenses, sophisticated air forces with significant air-to-air capacity, that would be a really stressful test, and they need to continue to advance the capabilities that were initially put into the F-35 to remain relevant." The APG-85 is, in essence, the specific system intended to make that advance possible. Without it, new Lot 17 and Lot 18 aircraft are technologically inferior to the very jets they were meant to replace.

Certification Time: The Operational Driver

Wittman, who has been conducting monthly phone calls with the F-35 Program Executive Officer — currently Marine Lt. Gen. Gregory Masiello, who assumed the role in July 2025 — identified radar certification time as a primary driver of the slippage. "The APG-81 could be certified in three days," he told reporters at the McAleese and Associates conference. "The newer radar takes much, much longer." Northrop Grumman has shortened the test cycle, he acknowledged, but not enough: "They're not doing that at the pace necessary for the aircraft coming off the line."

The certification requirement reflects genuine technical complexity. The APG-85's GaN-based transmit/receive modules — estimated at more than 2,400 individual elements — require extensive functional and electromagnetic compatibility validation before the radar can be cleared for flight on a stealth aircraft. The radar's LPI/LPD emission characteristics must be verified not to compromise the aircraft's signature. Its integration with the F-35's distributed sensor architecture, including the Distributed Aperture System and Electro-Optical Targeting System, must be validated through a multi-mode test regime. These are not bureaucratic delays; they are engineering imperatives. But planning them in parallel with an aggressive production schedule — rather than sequencing them appropriately — created the current crisis.

The Broader Block 4 Catastrophe

The APG-85 situation cannot be understood in isolation. It is one of multiple concurrent failures within the F-35's Block 4 modernization program — an effort that the Government Accountability Office, in its September 2025 report to Congress, characterized bluntly: "The F-35 program continues to overpromise and underdeliver."

Block 4 was originally conceived as a package of more than 66 discrete capability improvements, underpinned by a $1.9 billion hardware and software package called Technology Refresh 3 (TR-3), which provides the processing backbone — reportedly up to 37 times more computational throughput and 20 times more memory than its predecessor — needed to run advanced sensor fusion, weapons integration, and electronic warfare modes. TR-3, which was originally due in April 2023, has been beset by immature integrated core processor designs, Next Generation Distributed Aperture System delays, and software stability issues ranging from radar malfunctions to cockpit display failures. In 2024, operational testing of TR-3 software on the F-35 was reported as unlikely to commence before 2026.

As of the GAO's September 2025 assessment, Block 4 will not be complete until at least 2031 — at least five years behind the original schedule. The program has reduced the number of capabilities it will deliver, deferred those dependent on the Engine Core Upgrade until at least 2033, and eliminated others entirely as "no longer meeting warfighter needs." Block 4 costs had already grown from a $10.6 billion baseline to $16.5 billion as of 2021, with an updated cost estimate — expected in fall 2025 — still not publicly confirmed. The engine upgrade itself is now not expected to enter production until 2031, pushing full Block 4 capability with enhanced power and cooling to the mid-2030s at the earliest.

■ Delay Timeline: Key Events

• 2001 F-35 program baseline established; acquisition cost projected at $233 billion.
• 2012 Program rebaselined after costs breach statutory thresholds; estimate rises to $396 billion.
• Jan. 2023 Northrop Grumman officially announces APG-85 development contract.
• Apr. 2023 TR-3 software originally due; not delivered. Delivery halt follows.
• July 2023–July 2024 Year-long freeze on new F-35 deliveries due to TR-3 delays.
• Dec. 2024 Lot 18 contract awarded as undefinitized deal for 145 jets (83 U.S. military).
• Mar. 2025 Lockheed CEO Taiclet warns of APG-85 schedule risks in letter to USAF chief; proposes new dual-compatible bulkhead design.
• June 2025 F-35 deliveries with APG-85 bulkheads begin — without APG-85 radars.
• Sep. 2025 GAO: Block 4 delayed ≥5 years, $6B over budget; all 2024 deliveries averaged 238 days late.
• Sep. 2025 Lots 18–19 contract finalized; Lot 19 includes 69 U.S. military jets.
• Fall 2026 All new U.S. F-35 deliveries (A, B, C variants) expected without radar per Breaking Defense report.
• 2028 (est.) New dual-compatible bulkhead expected with Lot 20; APG-85 fielding may begin if program stays on revised schedule.

The Contractor Accountability Problem

Perhaps most damaging to program credibility is the incentive fee structure exposed by the GAO. Between 2021 and 2024, Lockheed Martin and Pratt & Whitney collected hundreds of millions of dollars in performance incentive fees even as delivery performance deteriorated. All 110 aircraft Lockheed delivered in 2024 arrived late — by an average of 238 days, nearly four times worse than the 61-day average lateness recorded in 2023. Pratt & Whitney delivered all 123 engines late in the same year. Yet contract structures permitted partial fee payments for jets delivered up to 60 days late, and when no aircraft met even that threshold, program officials redirected more than $100 million in unearned incentives to cover lab upgrades and TR-3 repairs rather than withholding them.

"Unless the F-35 program re-evaluates its use of incentive fees and better aligns them to achieving desired production schedule outcomes, it will be at greater risk of continuing to reward contractors for delivering engines and aircraft late," the GAO stated. The watchdog's conclusion is stark: the fee structure is "largely ineffective at holding the contractors accountable."

In this context, the decision to accept radar-less aircraft — allowing the production line to continue generating revenue for Lockheed Martin while the government absorbs the operational risk of non-combat-capable jets — is entirely consistent with the established pattern. The F-35 Joint Program Office noted that "the program in coordination with the Services deliberately undertook a highly concurrent development and production program" and that "this decision was made with full understanding of the risk." That may be accurate. The question is whether understanding a risk and appropriately managing it — including its financial implications for the contractor team — are the same thing.

Strategic Implications: A Shrinking Combat-Ready Fleet

The immediate tactical impact of radar-less deliveries is partially mitigated by the fact that, since July 2024, all new F-35 deliveries have already been non-combat-coded due to the unresolved TR-3 software situation. Combat operations — including strikes in the broader Middle East region — have continued with older TR-2 software jets from earlier production lots. In that sense, the addition of a radar gap to an already-existing software gap does not immediately reduce the number of combat-available aircraft.

The longer-term picture is considerably more troubling. As Pettyjohn noted, having to retrofit a large number of aircraft with the APG-85 at a later date "would be a huge blow to the tactical aircraft fleet, which is shrinking right now." Retrofit timelines and costs are presently unknown. The Air Force has already substantially cut its F-35 procurement plans, targeting just 39 aircraft in FY2027 and as few as 18 in FY2028 — reflecting a judgment that pre-Block 4 aircraft represent diminishing returns against peer threats. Aircraft without combat-capable software and without a radar compound that calculus. "Then we're starting to see where the numbers are going to belie the real capability that is available at any moment in time," Pettyjohn observed.

Foreign buyers of the F-35 are not affected. The APG-85 has not been cleared for export, and international customers continue to receive aircraft fitted with the APG-81. This divergence — where allied air forces are flying fully-equipped F-35s while U.S. forces take delivery of radar-less airframes — is a public relations and alliance management problem of the Pentagon's own creation.

The View From History: The Tornado F.2's 'Blue Circle Radar'

For any student of defense procurement, the F-35's radar-less deliveries carry an uncomfortable echo. In 1984, the Royal Air Force took delivery of the first production Panavia Tornado F.2 air defense variant — an aircraft designed specifically to intercept Soviet bombers, whose central capability was the GEC-Marconi AI.24 Foxhunter pulse-Doppler radar. The radar, several years late and 60 percent over budget, was not ready. The aircraft were delivered with concrete and lead ballast in the nose cones to maintain center-of-gravity balance.

RAF crews, in the tradition of British military dark humor, promptly named the missing sensor the "Blue Circle radar" — a play on the Blue Circle cement brand and the Rainbow Code naming convention used for British military avionics. The ballast was not merely an embarrassment. The first batch of 18 Tornado F.2s were restricted to training use, and the Foxhunter only entered service in 1985 in an interim standard that still did not meet requirements. A full capability standard was not achieved until the 1990s. The aircraft never fully exploited the capabilities of the AMRAAM and ASRAAM missiles it eventually carried.

The structural parallel is precise: a new airframe committed to a sensor that was not ready; an incompatible prior sensor that could not be substituted; ballast in the nose; training-only operations; unknown retrofit costs. Forty years of procurement reform, multiple generations of program management doctrine, the creation of the Office of the Director of Operational Test and Evaluation, the institutionalization of Milestone Decision Authority oversight — and the United States is in 2026 reproducing, at a scale 10 times larger and a cost several orders of magnitude greater, the same failure mode as the RAF in 1984.

"The bulkhead conversation should've been had years ago."

— Anonymous source familiar with the F-35 program, March 2026 (Breaking Defense)

The Path Forward — and Its Uncertainties

The near-term mitigation path is narrow. The F-35 Joint Program Office has stated that it has "plans to accelerate APG-85 production capacity," but details remain classified. In the best case, the radar reaches production maturity during the Lot 18 production run, limiting the number of ballast-nosed aircraft to a relative handful. In a more likely scenario, based on the program's historical trajectory, delays persist through Lot 19, potentially affecting the 69 U.S. military aircraft in that contract.

Lockheed Martin CEO Jim Taiclet acknowledged the risk formally in a March 2025 letter to then-Air Force Chief of Staff Gen. David Allvin, warning of schedule risks and committing to a redesigned forward fuselage that could accommodate either the APG-81 or APG-85. That dual-compatible bulkhead is not expected until Lot 20, whose deliveries begin in 2028. Even then, the redesign will require re-certification and integration testing — which itself could extend into the early 2030s for a full program-wide retrofit.

Wittman expressed confidence that the issue will eventually be resolved — but only eventually: "With the new radar, there's only so much that you can do to compress the time frame, just because of the rigor necessary in certifying a new radar system, a radar system that's much more capable, which the aircraft needs." That is a measured and technically accurate statement. It is also, from the perspective of operational commanders who need combat-ready aircraft today, cold comfort.

The APG-85 delay is, in the end, a symptom. The disease is a $2 trillion program whose acquisition strategy has for two decades prioritized industrial continuity and contractor revenue over the delivery of combat capability. The GAO has documented this for twenty years. Congress has been briefed. Corrective action requests have been issued. Incentive fees have been paid. And the jets, one by one, roll off the line with lead in their noses where their eyes should be.

Verified Sources & Formal Citations

1. Marrow, Michael; Insinna, Valerie; Stancy, Diana. "EXCLUSIVE: US poised to accept new F-35s without radars, sources say." Breaking Defense, March 19, 2026. Updated 17:00 ET.
   https://breakingdefense.com/2026/03/exclusive-us-poised-to-accept-new-f-35s-without-radars-sources-say/
2. Rogoway, Tyler; Trevithick, Joseph. "Are F-35s Being Delivered To The USAF Without Radars? (Updated)." The War Zone, February 11, 2026.
   https://www.twz.com/air/are-f-35s-being-delivered-to-the-usaf-without-radars-sure-seems-like-it
3. Rogoway, Tyler; Trevithick, Joseph. "Air Force Now Denies Receiving F-35s Without Radars." The War Zone, February 13, 2026.
   https://www.twz.com/air/air-force-now-denies-receiving-f-35s-without-radars
4. Cottle, Thomas. "Reports Suggest F-35s Are Being Delivered Without Radar Amid APG-85 Delays." The Aviationist, February 12, 2026.
   https://theaviationist.com/2026/02/12/reports-suggest-f-35s-delivered-without-radar/
5. Losey, Stephen. "Current F-35 Configuration Complicates Fielding Of APG-85 Radar." Aviation Today / Defense & Space, February 5, 2026.
   https://www.aviationtoday.com/2026/02/05/current-f-35-configuration-complicates-fielding-of-apg-85-radar/
6. U.S. Government Accountability Office. F-35 Joint Strike Fighter: Actions Needed to Address Late Deliveries and Improve Future Development. GAO-25-107632. Washington, D.C.: GAO, September 3, 2025.
   https://www.gao.gov/products/gao-25-107632  |  Full PDF: https://www.gao.gov/assets/gao-25-107632.pdf
7. Losey, Stephen. "Pentagon cuts back F-35 upgrades to slow schedule slips: Auditors." Defense News, September 3, 2025.
   https://www.defensenews.com/air/2025/09/03/pentagon-cuts-back-f-35-upgrades-to-slow-schedule-slips-auditors/
8. Pawlyk, Oriana. "F-35 Block 4 upgrade delayed until at least 2031: GAO." Breaking Defense, September 3, 2025.
   https://breakingdefense.com/2025/09/f-35-block-4-upgrade-delayed-until-at-least-2031-gao/
9. Tirpak, John A. "GAO: 'Action Needed' to Solve F-35 Block 4 Issues." Air & Space Forces Magazine, September 4, 2025.
   https://www.airandspaceforces.com/gao-action-needed-to-solve-f-35-block-4-issues/
10. Mehta, Aaron. "The jets were late. Lockheed got on-time bonuses anyway." Defense One, September 4, 2025.
    https://www.defenseone.com/business/2025/09/jets-were-late-lockheed-got-time-bonuses-anyway/407880/
11. "Why the U.S. Air Force Just Deeply Cut its F-35 Procurement Plans." Military Watch Magazine, November 12, 2025.
    https://militarywatchmagazine.com/article/why-usaf-deeply-cut-f35-plans
12. Rogoway, Tyler; Trevithick, Joseph. "F-35 Will Get New Radar Under Massive Upgrade Initiative." The War Zone, January 2023 (updated).
    https://www.twz.com/f-35-will-get-new-radar-under-massive-upgrade-initiative
13. "U.S. Delivering F-35As Without APG-85 Radar as Block 4 Upgrade Faces Integration Delays." The Defense News (independent), February 14, 2026.
    https://www.thedefensenews.com/news-details/US-Delivering-F-35As-Without-APG-85-Radar-as-Block-4-Upgrade-Faces-Integration-Delays/
14. "U.S. Delivers F-35 Lightning II Without AN/APG-85 Radar: $2 Trillion Stealth Program Faces Electromagnetic Warfare Setback." Defence Security Asia, February 11, 2026.
    https://defencesecurityasia.com/en/us-delivers-f35-without-apg85-radar-electromagnetic-warfare-delay/
15. "New US F-35s delivered with ballast in place of radar because of APG-85 delays." The Aviation Geek Club, February 13, 2026.
    https://theaviationgeekclub.com/new-us-f-35s-delivered-with-ballast-in-place-of-radar-because-of-apg-85-delays/
16. "Ballast Instead of Radar: Lockheed Martin Delivers F-35 Without AN/APG-85 Due to Delays." Militarnyi.com, February 2026.
    https://militarnyi.com/en/news/lockheed-martin-delivers-f-35-without-an-apg-85-due-to-delays/
17. "APG-81, APG-85: the five radar innovations of the F-35." War Wings Daily, January 30, 2026.
    https://warwingsdaily.com/apg-81-apg-85-the-five-radar-innovations-of-the-f-35/
18. Olson, Wyatt. "F-35 upgrade plagued by cost overruns and production delays, GAO report says." Stars and Stripes, September 3, 2025.
    https://www.stripes.com/theaters/us/2025-09-03/f35-fighter-modernization-gao-report-18969803.html
19. "AI.24 Foxhunter." Wikipedia / Wikimedia Foundation. Accessed March 2026.
    https://en.wikipedia.org/wiki/AI.24_Foxhunter
20. "Panavia Tornado ADV." Wikipedia / Wikimedia Foundation. Accessed March 2026.
    https://en.wikipedia.org/wiki/Panavia_Tornado_ADV
21. "Aviation Feature – the Tornado F3 Air Defence Variant (ADV)." Global Aviation Resource. Accessed March 2026.
    https://www.globalaviationresource.com/v2/2015/03/06/aviation-feature-the-tornado-f3-air-defence-variant-adv/
22. Northrop Grumman Corporation. Press Release: "Northrop Grumman to Develop AN/APG-85 Advanced Radar for F-35." January 11, 2023.
    [Official statement cited by The War Zone, source 12 above; direct Northrop press release no longer publicly indexed.]
23. F-35 Joint Program Office (JPO). Official statement to Breaking Defense, March 19, 2026 (post-publication update).
    [Cited in Breaking Defense exclusive, source 1 above.]

 

General Atomics Bets on Two Fronts: A Proliferating MQ-9B Fleet and the Dawn of Unmanned Air Combat

MQ-9 Drone Combat Power | General Atomics Interview - YouTube

With its MQ-9B SkyGuardian now anchoring European NATO ISR capacity across ten customer nations, and its YFQ-42A Collaborative Combat Aircraft advancing through pre-production flight testing, General Atomics Aeronautical Systems is executing the most aggressive dual-track expansion in the history of unmanned aviation.

March 19, 2026

Bottom Line Up Front

General Atomics Aeronautical Systems (GA-ASI) is simultaneously expanding its mature MQ-9B remotely piloted aircraft franchise across NATO Europe — with Germany's January 2026 procurement of eight SeaGuardians bringing the active customer list to ten nations — while advancing its YFQ-42A Collaborative Combat Aircraft through an accelerating flight-test program that has flown two production-representative airframes ahead of a fiscal year 2026 production decision.

The company is further extending the MQ-9B's mission envelope with a planned 2026 airborne early warning demonstration using Saab-supplied radar pods, long-range standoff weapons integration targeting JASSM, LRASM, and the Joint Strike Missile, and a short takeoff and landing configuration suitable for carrier operations. Taken together, these programs position GA-ASI to supply the backbone of both NATO theater ISR and emerging manned-unmanned teaming doctrine well into the 2030s.

FLORENNES AIR BASE, BELGIUM / SAN DIEGO — When the first Belgian Air Force MQ-9B SkyGuardian lifted off from Florennes Air Base on September 23, 2025, it completed a journey that had taken nearly seven years from the U.S. State Department's approval of the original sale to wheels-up on European soil. Following its arrival in August at Florennes, the aircraft made its maiden flight in Belgium on September 23, with 18 Belgian Air Force personnel having completed initial MQ-9B training at GA-ASI's Flight Test and Training Center in Grand Forks, North Dakota. The moment crystallized a strategic inflection point for GA-ASI: the San Diego-based unmanned aircraft manufacturer is no longer merely selling surveillance drones to European allies — it is fundamentally reshaping how NATO countries surveil their airspace, their maritime approaches, and increasingly, the Arctic.

The proliferation of MQ-9B in Europe delivers commonality between NATO countries. — Linden Blue, GA-ASI CEO

The scale of that reshaping has become clearer with each successive procurement announcement. On January 12, 2026, Germany's Federal Office of Bundeswehr Equipment, Information Technology and In-Service Support and the NATO Support and Procurement Agency (NSPA) announced the procurement of eight MQ-9B SeaGuardians from GA-ASI, with four Certifiable Ground Control Stations included and first delivery expected in 2028. Germany's decision brings the confirmed MQ-9B procurement list to ten countries. GA-ASI now holds MQ-9B procurement contracts with Belgium, Canada, Denmark, Germany, India, Japan, Poland, Taiwan, the United Kingdom, and the U.S. Air Force in support of Special Operations Command.

The NSPA has been instrumental in accelerating European uptake. In July 2025, Denmark and NSPA announced the procurement of four MQ-9B SkyGuardians including three Certified Ground Control Stations, with the platform selected for its pole-to-pole satellite control and de-icing capabilities to enable missions in the harsh conditions of the Arctic in support of Denmark and its NATO Allies. The NSPA framework creates a contractual structure that promotes interoperability while facilitating joint training across member nations — a deliberate architectural choice that effectively locks in commonality across an emerging NATO remotely piloted aircraft enterprise.

Certifiability as Strategic Differentiator

The MQ-9B's penetration of European markets rests on a design philosophy that distinguishes it sharply from its predecessor, the MQ-9A Reaper: the aircraft was engineered from the outset to operate in unsegregated civilian airspace. The MQ-9B SkyGuardian is designed to fly over the horizon via satellite for up to 40-plus hours in all types of weather and to safely integrate into civil airspace, meeting NATO STANAG 4671 airworthiness requirements with lightning protection, modified composite materials, and a detect-and-avoid system. That civil certifiability is not merely a marketing distinction — for European operators navigating congested continental airspace, it is an operational necessity.

The MQ-9B crossed a significant regulatory threshold in 2025. In 2025, MQ-9B became the first large remotely piloted aircraft to obtain a Military Type Certificate from the UK's Military Aviation Authority, certifying its safe operation without geographic restrictions, including over populous areas. The Royal Air Force's Protector RG Mk1 — the UK-designated variant of the MQ-9B — retired the service's MQ-9A Reaper fleet in September 2025 and has now assumed frontline duties. Ten of 16 Protectors had been delivered as of June 2025, with four in the UK and six in the United States used for testing and training.

MQ-9B Program Snapshot — March 2026

Flight Hours Accumulated	9 million+ (entire Predator lineage)
Max Endurance	40+ hours (SkyGuardian); 30+ hours (SeaGuardian)
Wingspan	79 ft (24 m)
Max External Payload	4,750 lb (2,155 kg); nine hardpoints
Operating Cost (MQ-9B)	Below $5,000/flight hour (GA-ASI estimate)
NATO STANAG Compliance	4671 (civil airspace certifiable)
Confirmed Customer Nations	Belgium, Canada, Denmark, Germany, India, Japan, Poland, Taiwan, UK, USA
UK MTC	Awarded 2025 (first large RPA to receive)
Long-range Weapons Test	Planned 2026 (JASSM, LRASM, JSM integration)
AEW Demo	Planned summer 2026, Desert Horizon, CA

Mission Expansion: From Surveillance to Strike to Early Warning

GA-ASI has systematically broadened the MQ-9B's mission aperture beyond the intelligence, surveillance, reconnaissance, and targeted strike roles that defined the original Predator lineage. The most consequential near-term expansion is a planned integration of long-range standoff weapons. In February 2026, GA-ASI announced it is developing the addition of long-range standoff weapons to the MQ-9B SkyGuardian and SeaGuardian, with engineers adapting the aircraft's payload, stability, range, and other features to accommodate the new generation of extended-range precision weapons. The company is examining integration of the Lockheed Martin Joint Air-to-Surface Standoff Missile, the Long-Range Anti-Ship Missile, and the Kongsberg/Raytheon Joint Strike Missile, and plans to flight-test at least one long-range weapon as early as 2026.

The anti-submarine warfare mission has similarly expanded. SeaGuardian integrates a centerline wide-area maritime radar, an automatic identification system, electronic support measures, and a self-contained anti-submarine warfare mission kit — and is the first remotely piloted aircraft in its class to enable real-time search and patrol above and below the ocean's surface. With four wing stations available for sonobuoy dispensing pods, a configured SeaGuardian can deploy up to 80 sonobuoys in a single sortie and conduct submarine prosecution across a mission radius of 1,200 nautical miles — an asymmetric capability against peer-state undersea threats in the Baltic and North Atlantic.

AEW is the ideal mission for an unmanned, semi-autonomous vehicle. We can do 16 to 20 hours of endurance at much lower cost than manned platforms. — Satish Krishnan, GA-ASI VP, MQ-9B International

The most technically ambitious near-term mission extension is airborne early warning. In June 2025, GA-ASI announced a partnership with Saab to develop airborne early warning and control capability for the MQ-9B line, with plans to fly the AEW configuration in 2026. Following their Paris Air Show announcement, GA-ASI and Saab confirmed in November 2025 at the Dubai Airshow that the AEW demonstration will be conducted in the summer of 2026 at GA-ASI's Desert Horizon flight operations facility in Southern California. The system will carry Saab-supplied radar pods — derived from the company's Erieye ER gallium-nitride active electronically scanned array technology — under each wing, with a centerline pod housing data processing equipment.

The UK Royal Navy's Carrier Strike Airborne Early Warning requirement has emerged as a major potential customer, with UK defence procurement minister Maria Eagle confirming in May 2025 that the MQ-9 was being considered as a candidate to succeed the Crowsnest system when it reaches end of life. The carrier application requires the MQ-9B STOL configuration, which was demonstrated aboard HMS Prince of Wales in November 2023. The partners have stated that the AEW capability, with service entry offered before the end of this decade, will provide persistent surveillance across a wide range of applications including early detection and warning, long-range target tracking, and flexible combat system integration over both line-of-sight and SATCOM connectivity.

The Collaborative Combat Aircraft: Mass at Speed

GA-ASI's second major thrust is, by the company's own characterization, its top strategic priority for this period: the YFQ-42A Collaborative Combat Aircraft. The program represents a categorical departure from the MALE surveillance role that defined the company's first three decades. Where the MQ-9B is optimized for endurance and sensor integration, the YFQ-42A is a jet-powered air combat platform designed to operate as a semi-autonomous wingman to crewed fighters.

The U.S. Air Force selected GA-ASI in 2024 to develop and build the YFQ-42A, with the designation announced in March 2025. The "Y" prefix indicates a production-representative prototype — it will be dropped upon entering production — while "F" signifies a fighter aircraft and "Q" denotes an uncrewed system. The YFQ-42A features an elongated fuselage with slender wings, a dorsal-mounted inlet, a single engine, V-tails, and an internal weapons bay planned to carry two AIM-120 AMRAAMs. The design is derived from the company's XQ-67A Off-Board Sensing Station demonstrator developed for the Air Force Research Laboratory, with modifications for greater speed and fighter-like maneuverability.

On August 27, 2025, GA-ASI announced the YFQ-42A had begun flight testing in coordination with the U.S. Air Force. The program had completed ground testing beginning in May 2025, achieving first flight in just over a year from program selection — a development tempo that GA-ASI President David R. Alexander described as an "incredible achievement." By November 2025, GA-ASI had flown a second YFQ-42A, with a photo showing two aircraft on the flightline with different markings and numbering from the first prototype — stepping up the pace of testing for the new uncrewed fighter escort.

YFQ-42A CCA Program Snapshot — March 2026

Program	USAF Increment I Collaborative Combat Aircraft
Competing Design	Anduril YFQ-44A "Fury"
Lineage	XQ-67A Off-Board Sensing Station / Gambit family
Configuration	Single engine, dorsal inlet, V-tail, internal weapons bay
Planned Armament	2× AIM-120 AMRAAM (internal)
Combat Radius	>700 nautical miles (USAF estimate)
RCS Profile	Reduced signature (~F-35 class, per USAF graphics)
Ground Testing	May 2025
First Flight	August 27, 2025
Second Aircraft Flown	November 2025
Production Decision	Expected FY2026
USAF Acquisition Goal	1,000+ CCAs
European CCA Variant	Announced July 2025; GA-ATS manufacturing in Oberpfaffenhofen

The production calculus is central to the CCA concept. The GA-ASI program has focused on creating a high-rate production environment that enables the Air Force to reach its goal of producing more than 1,000 CCAs on an accelerated timeline. GA-ASI manufactures more than 100 aircraft per year at its five million-square-foot facility in Poway, California — an industrial baseline that few competitors can match. A competitive Increment I production decision is expected in fiscal year 2026, with the Air Force previously indicating between 100 and 150 Increment I CCAs could be acquired, though it remains unclear whether that fleet will consist entirely of YFQ-42As, YFQ-44As, or a mix of both types.

GA-ASI has simultaneously launched a European CCA derivative. Announced at the Royal International Air Tattoo in July 2025, the European CCA will be derived from the YFQ-42A baseline with European mission system customization and manufacturing supported by General Atomics' German aerospace affiliate, General Atomics Aerotec Systems GmbH, headquartered in Oberpfaffenhofen near Munich. The initiative leverages the transatlantic industrial relationship between GA-ASI and GA-ATS to offer European NATO air forces a path to domestic production of a mature unmanned combat platform without the years of development risk involved in a clean-sheet program.

Architecture for Contested Airspace

The strategic coherence linking the MQ-9B expansion and the YFQ-42A program lies in a shared analytical conclusion: that future NATO air operations in contested environments will require mass, persistence, and expendability that no crewed platform fleet can economically provide. The MQ-9B addresses the persistence requirement — its operating cost of below $5,000 per flight hour makes it a credible complement to far more expensive airborne early warning and maritime patrol platforms. The YFQ-42A addresses mass and expendability in the fighter escort role, designed to absorb the attrition risks that would otherwise fall on irreplaceable crewed aircraft and pilots.

Innovative software systems, for example, allow a single human controller to operate multiple MQ-9B aircraft at once — a flight of MQ-9Bs fanning out over a large section of territory or ocean and patrolling semi-independently under the supervision of an operator working via satellite link. GA-ASI's Quadratix software ecosystem provides the autonomy, machine learning, and AI layer that enables reduced crew requirements for the AEW mission package as well. The operator workload reduction implied by these software capabilities is not incidental — it is architecturally necessary if NATO air forces operating with constrained personnel budgets are to extract full value from enlarged unmanned fleets.

The intersection of these programs with NATO's post-2022 European security posture is direct. NATO has added more than 800 miles of frontier along its eastern front since Finland and Sweden acceded to the Alliance, and northern European nations that once stood apart from active defense planning have begun moving swiftly to address new security imperatives. The Arctic surveillance gap, the Baltic maritime threat environment, and the demand for persistent wide-area ISR at the NATO eastern flank all define requirements the MQ-9B was purpose-built to meet. Belgium's operational performance with the SkyGuardian, which has been noted with satisfaction by GA-ASI executives, serves as a proof-of-concept for other NATO operators watching closely.

Whether both YFQ-42A and YFQ-44A will survive the upcoming Increment I production competition remains uncertain. The Air Force is taking a rapid and flexible approach with the CCA program, with at least one of the Increment I prototypes advancing to production — but the service intends to allow other firms besides GA-ASI and Anduril to also compete for the full production contract. What is certain is that General Atomics enters that competition having already demonstrated two production-representative airframes in flight — a head start that, in a program measured in months rather than years, is not easily overcome.

---

Verified Sources & Formal Citations

[1] General Atomics Aeronautical Systems, Inc. "MQ-9B SeaGuardian Product Page." GA-ASI Official Website. https://www.ga-asi.com/remotely-piloted-aircraft/mq-9b-seaguardian

[2] General Atomics Aeronautical Systems, Inc. "Germany Buys Eight MQ-9B SeaGuardian RPA Through NSPA." Press Release, January 12, 2026. https://www.ga.com/germany-buys-eight-mq-9b-seaguardian-rpa-through-nspa

[3] General Atomics Aeronautical Systems, Inc. "MQ-9B SkyGuardian Product Page." GA-ASI Official Website. https://www.ga-asi.com/remotely-piloted-aircraft/mq-9b-skyguardian

[4] General Atomics Aeronautical Systems, Inc. "Denmark Buys Four MQ-9B SkyGuardians from GA-ASI." Press Release, July 23, 2025. https://www.ga-asi.com/denmark-buys-four-mq-9b-skyguardians-from-ga-asi

[5] General Atomics Aeronautical Systems, Inc. "MQ-9B SkyGuardian Flies for First Time in Belgium." Press Release, September 24, 2025. https://www.ga.com/mq-9b-skyguardian-flies-for-first-time-in-belgium

[6] Naval Technology. "GA-ASI to Arm MQ-9B SeaGuardian, SkyGuardian with Long-Range Weapons." Published March 2026. https://www.naval-technology.com/news/ga-asi-mq9b-seaguardian-weapons/

[7] General Atomics Aeronautical Systems, Inc. "GA-ASI Develops Long-Range Weapons Capabilities for MQ-9B." Press Release, February 23, 2026. https://www.ga.com/ga-asi-develops-long-range-weapons-capabilities-for-mq-9b

[8] Wikipedia. "General Atomics MQ-9 Reaper." Last modified March 2026. https://en.wikipedia.org/wiki/General_Atomics_MQ-9_Reaper

[9] General Atomics Aeronautical Systems, Inc. "MQ-9B: Securing Northern Europe." GA-ASI Official Website. https://www.ga-asi.com/mq-9b-securing-northern-europe

[10] General Atomics Aeronautical Systems, Inc. "GA-ASI Adds Saab Airborne Early Warning Capability to MQ-9B." Press Release, June 15, 2025. https://www.ga-asi.com/ga-asi-adds-saab-airborne-early-warning-capability-to-mq-9b

[11] General Atomics Aeronautical Systems, Inc. "GA-ASI and Saab Will Demonstrate AEW&C on MQ-9B in 2026." Press Release, November 17, 2025 (Dubai Airshow). https://www.ga.com/ga-asi-and-saab-will-demonstrate-aew-c-on-mq-9b-in-2026

[12] Naval News. "General Atomics Brings Saab Onboard for MQ-9B AEW Mission Package." June 19, 2025. https://www.navalnews.com/naval-news/2025/06/general-atomics-brings-saab-onboard-for-mq-9b-aew-mission-package/

[13] FlightGlobal. "General Atomics, Saab Cite 'Tremendous Customer Interest' in MQ-9B AEW Development." November 25, 2025. https://www.flightglobal.com/general-atomics-saab-cite-tremendous-customer-interest-in-mq-9b-aew-development/165478.article

[14] Janes. "Paris Air Show 2025: General Atomics and Saab Partner on AEW MQ-9B." June 2025. https://www.janes.com/osint-insights/defence-news/air/paris-air-show-2025-general-atomics-and-saab-partner-on-aew-mq-9b

[15] The Aviationist. "General Atomics and Saab Join Forces for MQ-9B Airborne Early Warning Capability." June 16, 2025. https://theaviationist.com/2025/06/16/ga-asi-saab-mq-9b-airborne-early-warning/

[16] General Atomics Aeronautical Systems, Inc. "GA-ASI Welcomes USAF Designation for New CCA: YFQ-42A." Press Release, March 3, 2025. https://www.ga-asi.com/ga-asi-welcomes-usaf-designation-for-new-cca-yfq-42a

[17] General Atomics Aeronautical Systems, Inc. "GA-ASI Marks Another Aviation First With YFQ-42A CCA Flight Testing." Press Release, August 27, 2025. https://www.ga.com/ga-asi-marks-another-aviation-first-with-yfq-42a-cca-flight-testing

[18] U.S. Air Force. "Collaborative Combat Aircraft, YFQ-42A Takes to the Air for Flight Testing." August 27, 2025. https://www.af.mil/News/Article-Display/Article/4287627/collaborative-combat-aircraft-yfq-42a-takes-to-the-air-for-flight-testing/

[19] Air & Space Forces Magazine. "General Atomics Flies Second CCA, Debuts Ground-Attack Drone." John A. Tirpak, November 4, 2025. https://www.airandspaceforces.com/general-atomics-flies-second-cca-ground-attack-gambit/

[20] The War Zone. "Our First Look At The YFQ-42 'Fighter Drone' Collaborative Combat Aircraft." May 19, 2025. https://www.twz.com/air/our-first-look-at-the-yfq-42-fighter-drone-collaborative-combat-aircraft

[21] The Aviationist. "General Atomics YFQ-42A Collaborative Combat Aircraft Officially Revealed." May 19, 2025. https://theaviationist.com/2025/05/19/general-atomics-yfq-42a-revealed/

[22] DefenseScoop. "General Atomics Begins Flight Tests for Air Force CCA Drone Program." August 27, 2025. https://defensescoop.com/2025/08/27/general-atomics-cca-begins-flight-tests-air-force-drone-program/

[23] General Atomics. "A New Transatlantic Partnership for European CCA." Press Release, July 17, 2025. https://www.ga.com/a-new-transatlantic-partnership-for-european-cca

[24] Wikipedia. "General Atomics YFQ-42 Dark Merlin." Last modified February 2026. https://en.wikipedia.org/wiki/General_Atomics_YFQ-42

[25] Breaking Defense. "General Atomics' MQ-9B Ready for Game-Changing AEW Capability to Strengthen European Defense." June 16, 2025. https://breakingdefense.com/2025/06/general-atomics-mq-9b-ready-for-game-changing-aew-capability-to-strengthen-european-defense/

Aviation Week Staff Report  |  March 19, 2026  |  All rights reserved. Research compilation based on official press releases, industry publications, and open-source defense reporting.

 

 

The Illusion of Knowledge: How Archaeology Has Been Reading the Wrong Pages

 The Journal of Record for Field Archaeology and Remote Sensing

BOTTOM LINE UP FRONT: 

Conventional archaeology has constructed its understanding of human prehistory from a small and geographically biased sample. Conservative estimates suggest that the vast majority of the human past lies hidden beneath tropical forest canopies, desert sands, and the post-glacial ocean — three environments in which systematic survey has barely begun. New remote sensing technologies — airborne LiDAR, Synthetic Aperture Radar, bathymetric sonar, and AI-enhanced satellite analysis — are dismantling this illusion with each passing field season, repeatedly revealing that populations were larger, civilizations more sophisticated, and human landscapes more extensively engineered than the terrestrial excavation record suggested. Simultaneously, armed conflict, deliberate heritage destruction, looting for antiquities revenue, and underfunding in regions critical to understanding the full human story are actively erasing what survives before it can be documented. What we thought we knew about the past is an artifact of where we chose to look — and that is changing.

■ Cover Feature

Tropical jungle, desert sand, and ocean shelf have hidden the majority of the human past. A new generation of sensors — and a reckoning about where the gaps lie — is forcing a fundamental revision of what we think we know.

Special Report  ·  March 2026  ·  Archaeology Today Investigative Desk  ·  With research contributions from Correspondent Teams in the Americas, Southeast Asia, the Middle East, and Sub-Saharan Africa

• 27mkm² of continental shelf exposed at the last glacial maximum — now submerged
• ~80%estimated share of human past unexamined by systematic survey
• 2,600+submerged prehistoric sites catalogued across 19 countries by splashcos

In the spring of 2015, a helicopter carrying a laser scanning instrument flew for twenty hours over the jungles of northwest Cambodia. When archaeologist Damian Evans processed the resulting data — filtering out twelve centuries of regrown forest — he found himself staring at something that should not have existed: an entire city, known from stone inscriptions as Mahendraparvata, perfectly preserved in the topography of Phnom Kulen mountain. "We have entire cities discovered beneath the forest that no one knew were there," Evans later told reporters. "A city that we figured wasn't there just appeared." He had spent a decade walking the ground in search of it.

That moment encapsulates one of the most significant epistemological crises in modern science. Archaeology — the discipline charged with reconstructing the full human story — has been conducting its investigations with instruments and methods that are essentially blind to three of the most important environments in which that story unfolded: dense tropical forest, deep desert sand, and the ocean floor. The result is a record that is not merely incomplete but systematically distorted, weighted toward the open, arid, or temperate landscapes where traditional fieldwork is convenient.

The reckoning is now underway. A convergence of remote sensing technologies — airborne LiDAR (Light Detection and Ranging), Synthetic Aperture Radar (SAR), multibeam sonar, sub-bottom profiling, and AI-enhanced satellite analysis — is stripping away the obscuring layers faster than at any previous point in the history of the discipline. Every major survey of the past decade has arrived at the same conclusion: the populations were larger, the landscapes more engineered, and the civilizations more sophisticated than the conventional record suggested. The wilderness was not wild.

How LiDAR Works

A LiDAR system fires laser pulses at rates of up to one million per second from an aircraft. When pulses encounter dense forest canopy, most reflect from leaves. But a statistical fraction penetrate gaps and strike the ground. By computing the precise return time of every pulse, the system builds a three-dimensional elevation model of the ground surface below. Processing software then strips the vegetation layer, leaving a bare-earth terrain model that reveals every mound, platform, road cut, and terrace that human hands created — features invisible to ground observers and satellite photography alike.

Vertical accuracy: ±4–6 inches at ground level. Coverage rate: Hundreds of square kilometers per flight day. A survey that would take decades on foot can be completed in weeks from the air.

Synthetic Aperture Radar (SAR) operates on a complementary principle for desert environments: radar wavelengths penetrate dry sand to depths of several meters, reflecting off buried structures — walls, roads, foundations — that are invisible at the surface.

The Three Hidden Domains

Tropical Forest: Dense canopy blocks optical and satellite observation of ground-level features. Traditional survey requires machetes and years. LiDAR penetrates the canopy.

Desert Sand: Wind-blown overburden buries structures under meters of sediment within centuries. SAR radar penetrates dry sand to reveal foundations and road networks.

Continental Shelf: Sea levels were 120–130 meters lower at the Last Glacial Maximum, exposing vast inhabited plains now submerged. Bathymetric sonar and sub-bottom profiling map these drowned landscapes.

The Jungle Record

What the Forest Was Hiding: The Americas and Southeast Asia

The scale of what LiDAR has revealed in the tropical Americas in the past decade defies straightforward summary. The single most consequential survey was the 2016 PACUNAM LiDAR Initiative in Guatemala's Petén jungle, which mapped 2,144 square kilometers of the Maya Biosphere Reserve and found 61,480 previously unknown structures — isolated farmsteads, fortresses, causeways, palaces, and field systems — invisible to a century of prior ground survey. Population estimates for the Maya Classic period had to be revised upward from roughly five million to as many as ten to fifteen million. What had appeared to be scattered city-states in a wilderness turned out to be an interconnected urban civilization that modified virtually every square kilometer of its landscape.

The cascade of discovery has continued. In October 2024, researchers formally named Valeriana, a Maya city in Mexico's Campeche state that had been lurking in LiDAR data since 2013 — over forty square miles of pyramids, causeways, ball courts, and amphitheaters, the second-largest Maya site yet identified. A 2023 Tulane University study used PACUNAM LiDAR data to identify 110,000 structures across the central Maya lowlands, and demonstrated that 30 percent featured elite masonry construction — providing, for the first time, a civilization-scale map of wealth distribution in an ancient society.

In Ecuador's Upano Valley, French researcher Stéphen Rostain spent nearly thirty years knowing something significant lay beneath the forest, but lacking the overview to understand it. A 2015 LiDAR survey changed that. The resulting data, published in Science in January 2024, revealed more than 6,000 earthen platforms, five major cities, and a road network that crossed at right angles — not following the landscape but overriding it — across 300 square kilometers. The settlements, occupied between roughly 500 BCE and 600 CE, are more than a millennium older than any other known complex Amazonian society. "It changes the way we see Amazonian cultures," said co-author Antoine Dorison. "Most people picture small groups, probably naked, living in huts and clearing land — this shows ancient people lived in complicated urban societies."

Bolivia's Llanos de Mojos yielded comparably startling results in a 2022 Nature paper. LiDAR surveys of the Casarabe culture's territory revealed two sites of 147 and 315 hectares respectively in a four-tiered settlement system — complete with conical pyramids up to 22 meters tall, U-shaped platform mounds, curved moats, and a network of straight causeways radiating outward in all directions. As Roland Fletcher of the University of Sydney observed in response: the Congo Basin, equally forested and equally unsurveyed, probably hosted comparable low-density urban development.

"The very biodiversity we seek to safeguard may itself be a legacy of centuries or millennia of human intervention."

— Prof. José Iriarte, University of Exeter, on Amazon landscape archaeology

The implications extend beyond archaeology. Research synthesized in Archaeology of the Amazon (2024) documents thirteen millennia of continuous human occupation across a territory larger than the European Union, demonstrating that the Amazon's extraordinary biodiversity is not pristine wilderness but the managed legacy of populations who selectively cultivated useful species across millennia. The "virgin forest" is a colonial myth. This finding has direct relevance to contemporary conservation policy, which has for decades treated the Amazon as a naturally unmodified baseline.

In Southeast Asia, the Cambodian Archaeological LiDAR Initiative's 2012 and 2015 surveys produced results equally transformative. The Khmer Empire, long understood primarily through its stone temples, was revealed as a civilization of low-density urban sprawl extending over 1,000 square kilometers — the largest preindustrial city complex in the world. The LiDAR data also provided a key to understanding the empire's collapse: severe erosion patterns in the hydraulic infrastructure suggest that a complex water management system, strained by the same climate shifts that ended the Medieval Warm Period, failed catastrophically in the 14th and 15th centuries.

The Desert Record

Cities Under Sand: Arabia, the Sahara, and the Illusion of Emptiness

Desert archaeology has a different technical challenge. Where LiDAR is the primary instrument in the tropics, SAR radar and AI-enhanced satellite analysis dominate desert work. Dry sand is, paradoxically, an excellent medium for radar penetration: wavelengths that would be absorbed by moisture in humid soil pass through the desiccated overburden and reflect off buried stone, ceramic, and compressed-mud structures below.

The Garamantes civilization of Libya is perhaps the most thoroughly documented case of a desert society rescued from near-invisibility by remote sensing. Satellite photography and radar surveys led by David Mattingly of the University of Leicester have, across 2,500 square kilometers of the Fezzan region of southern Libya, identified 158 major settlements, 184 cemeteries, 30 square kilometers of agricultural fields, and extensive underground irrigation systems (foggaras) — networks of tunnels tapping fossil groundwater that created green oases in what is now hyperarid desert. The Garamantes, who ruled from roughly 1000 BCE to 700 CE, built a sophisticated state with walled towns, trade networks extending to Rome, and an engineering tradition that sustained large populations in an environment that today supports almost none.

Recent AI-enhanced SAR work in the Arabian Peninsula's Empty Quarter (Rub' al-Khali) has identified traces of 5,000-year-old settlements and trade-route infrastructure — a finding that overturns the assumption that this vast expanse was always the uninhabited wasteland it appears today. During the early Holocene, when monsoon patterns delivered substantially more rainfall to Arabia, the peninsula was partially vegetated and supported complex communities. Their material traces are buried under shifting dunes and accessible only to penetrating radar.

The Endangered Archaeology in the Middle East and North Africa (EAMENA) project, coordinated from the universities of Oxford, Leicester, and Durham, has used satellite imagery, LiDAR where available, and machine learning to systematically document archaeological sites across twenty countries — a project with acute urgency given the destruction rate in conflict zones. As the 2025 Annual Review of Anthropology noted, AI integration with remote sensing is now allowing researchers to identify potential sites with up to 80 percent accuracy from satellite signatures alone, transforming the economics of survey work in inaccessible regions.

The Submerged Record

The World Beneath the Waves: Drowned Landscapes of the Continental Shelf

The least surveyed of all three hidden domains is the most significant in terms of sheer temporal depth. During the Last Glacial Maximum (approximately 26,000–19,000 years ago), sea levels stood 120–130 meters lower than today. Approximately 27 million square kilometers of continental shelf — an area larger than Africa — was exposed dry land, inhabited by the ancestors of every modern human population and then drowned as the ice melted. Every coastal archaeological site in the world is the wrong site. The real coastal sites, where humans preferentially concentrated, are underwater.

The SPLASHCOS network (Submerged Prehistoric Archaeology and Landscapes of the Continental Shelf), funded by the European Commission as COST Action TD0902 from 2009 to 2013, provided the first systematic international inventory, cataloguing over 2,600 submerged prehistoric sites across 19 countries. The successor ArCHe doctoral network, funded under EU Horizon, is training a new generation of researchers specifically in submerged landscape methodology. As SPLASHCOS researchers wrote in Antiquity, existing syntheses of world prehistory "are likely to be seriously incomplete" because most of the great transformations of human history — the dispersal of modern humans, the origins of seafaring, the roots of early farming economies — took place when sea levels were lower.

The most active survey program is in the North Sea, where University of Bradford's Submerged Landscapes Research Centre, led by Professor Vincent Gaffney, has now mapped 188,000 square kilometers of the drowned plain known as Doggerland — once a landmass the size of England and Scotland combined connecting Britain to continental Europe. A 2025 release from Bradford announced the deployment of AI to analyze the completed bathymetric map and identify "Goldilocks Zones": areas where the right combination of habitable terrain, preservation potential, and accessibility make archaeological finds most likely. Sediment cores collected in 2024 are being analyzed for ancient environmental DNA.

Active offshore wind development across the North Sea shelf has given the project unusual institutional support: developers are legally required to commission archaeological impact assessments before installation, producing survey data that feeds directly into research. A 2024 report from the University of Bradford noted that this partnership with industry has become a primary driver of new data acquisition across Doggerland's extent.

·       1931 Barbed antler harpoon dredged from North Sea — first direct artifact from Doggerland human habitation

·       1984 Atlit-Yam discovered off Israel: 9,000-year-old submerged Neolithic village with stone wells, megaliths, and 90+ burials

·       2009–13 SPLASHCOS network catalogues 2,600+ submerged sites across 19 countries

·       2015–21 ERC-funded Lost Frontiers project maps Doggerland; extracts aDNA from submerged river valley cores

·       2024–25 Bradford's AI-assisted Goldilocks analysis launched; 2024 sediment core collection underway. ArCHe doctoral network begins

·       2024–29 MAEASaM Phase 2 expands submerged and buried site mapping to DRC, Mozambique, The Gambia

The African Gap

The Most Important Blank on the Map: Sub-Saharan Africa and the Congo

No omission in the global archaeological record is more consequential — or more clearly the product of structural inequity rather than genuine absence — than sub-Saharan Africa. A 2024 systematic review found that Africa south of the Sahara is represented by only two LiDAR archaeological projects in the entire peer-reviewed literature. The Congo Basin, one of the world's largest tropical forests and the heartland of the Bantu expansion that populated the southern half of the continent, has received essentially zero systematic airborne LiDAR survey. As a 2022 scholarly review put it bluntly: "Although archaeological survey using remote sensing imagery is common in many parts of the world, its use in sub-Saharan Africa remains rare."

This is not because there is nothing to find. Africa's known pre-colonial urban civilization includes Great Zimbabwe (a stone-built capital city of approximately 18,000 people, dating from the 11th century, with over 400 associated stone settlements scattered across the southern African plateau); the Kingdom of Kongo (with its capital Mbanza Kongo estimated at 100,000 inhabitants — comparable to contemporary Lisbon — at its 16th-century peak); the Yoruba city-state network of Nigeria; the Benin Empire with its sophisticated bronze-casting tradition; and the Swahili Coast trading cities from Kilwa to Mombasa. None of these have been subjected to systematic LiDAR survey of their surrounding hinterlands. The full extent of their urban landscapes, field systems, and road networks is simply unknown.

Roland Fletcher of the University of Sydney, whose work on low-density urbanism encompasses Angkor, the Casarabe culture, and the Maya, has explicitly predicted that the Congo Basin, given its ecological and demographic parallels with the surveyed tropical forests, "probably hosted other early forms of low-density urban development." The prediction is grounded in the same reasoning that proved correct for Bolivia, Ecuador, Guatemala, and Cambodia.

The structural obstacles are real: vast and poorly-roaded terrain, decades of armed conflict in the DRC and neighboring states, inadequate research infrastructure, and the chronic underfunding of African archaeology relative to its global importance. The MAEASaM (Mapping Africa's Endangered Archaeological Sites and Monuments) project completed Phase 1 in June 2024, having documented thousands of sites across eight countries using multispectral satellite imagery and automated detection. Phase 2 (2024–29), funded through the Arcadia Foundation, has expanded coverage to include the DRC. It is the first project to bring systematic remote sensing to the Congo at scale — a starting point, but a significant one.

"The blank on the LiDAR map of Africa is not evidence of absence. It is, almost certainly, evidence of absence of surveying."

— Analysis consensus, Archaeology Today, 2026

The Destruction Problem

While We Look the Other Way: War, Looting, and the Permanent Loss of the Record

If the technology revolution represents the discipline's greatest opportunity in a century, the destruction of the archaeological record in conflict zones represents its gravest contemporary crisis. The two trends are running in parallel, and in the most critical regions they are running in opposite directions.

A peer-reviewed satellite analysis published in Near Eastern Archaeology found that more than 25 percent of archaeological sites in Syria have been impacted by looting since the war began — an order-of-magnitude increase over pre-war rates. At Mari, one of the world's most important Bronze Age sites on the Euphrates, archaeologist Pascal Butterlin of Pantheon-Sorbonne documented near-total destruction of the site's central palace complex. At Dura-Europos — until recently one of the best-preserved Roman-Persian frontier cities in the world — satellite imagery recorded approximately 9,500 looting pits in the site's residential and necropolis areas.

UN Security Council Resolution 2347 (2017) was the first in history to explicitly frame the deliberate destruction of cultural heritage as a security imperative, and designated its intentional targeting as a war crime. A 2024 analysis in the International Journal of Cultural Property found that the resolution's implementation has been starkly selective: while it has been applied vigorously to ISIS-related destruction in Iraq and Syria, comparable heritage destruction in Yemen (by Saudi airstrikes), in Gaza, by Boko Haram in Nigeria, and in Myanmar has not been subject to equivalent censure. The authors conclude that the securitization of cultural heritage, while creating useful legal tools, has been deployed with political inconsistency.

UNESCO's 2015 report on cultural property and armed conflict documented that "blood antiquities" may have constituted up to fifteen to twenty percent of ISIS revenue at peak, making antiquities trafficking one of the organization's primary financing mechanisms — funded, in part, by collectors and auction houses in Europe, the United States, and the Gulf states. A 2019 prosecution in the Netherlands and a 2021 civil forfeiture case in the United States established legal precedents for holding downstream purchasers liable, but enforcement remains scattered.

In Afghanistan, where continuous conflict has prevented systematic archaeological work since 1979, a 2018 satellite analysis found that looting and destruction patterns during Taliban periods and following the 2001 US invasion were effectively continuous — the public narrative of ISIS as the primary heritage threat masked decades of pre-existing damage. Yemen, another archaeologically critical region containing some of the world's earliest evidence for settled agriculture and iron-age urbanism, has been effectively closed to international archaeology since 2015. The chair of Yemen's General Organization of Antiquities described the situation to Science magazine in direct terms: "Our immortal history has been wasted by wars."

Ukraine presents the most recent high-profile case. Following Russia's 2022 invasion, World Heritage Committee deliberations in 2024 over the wording of conservation reports concerning targeted destruction in Odessa, Kyiv, and Lviv — all UNESCO World Heritage cities — revealed the extent to which diplomatic considerations constrain even the naming of perpetrators in formal heritage protection processes.

The Path Forward

A New Archaeology: What the Next Decade Requires

Technology

Drone-mounted LiDAR systems are rapidly reducing the cost and logistical barriers to airborne survey. Systems that required fixed-wing aircraft or helicopters a decade ago can now be mounted on commercial UAVs capable of operating from road-accessible field camps. The 2024 Annual Review of Anthropology noted that AI-assisted feature detection is now achieving 80–95 percent accuracy on known site types, dramatically reducing the analytical bottleneck that previously slowed the conversion of raw LiDAR data into publishable results. Bathymetric LiDAR — which works in water as well as on land — is extending coverage to shallow coastal zones previously inaccessible to ship-based sonar.

Funding & Access

The structural funding gap between African, Southeast Asian, and Pacific Island archaeology and the Americas/Europe must be addressed at the institutional level. The MAEASaM project's model — combining Cambridge-based coordination with rigorous in-country training partnerships — has demonstrated what systematic survey can achieve even with constrained resources. The offshore wind and extractive industries, which have legal obligations to commission archaeological surveys before ground disturbance, represent a largely untapped source of survey data and funding that the discipline has only recently begun to systematically exploit.

Heritage Protection

The legal architecture developed since 2015 — UNSC Resolution 2347, national-level cultural property protection legislation in Council of Europe member states, the US civil forfeiture framework for looted antiquities — requires consistent application. The political inconsistency documented in the International Journal of Cultural Property analysis (2024) undermines both the enforcement deterrent and the legitimacy of international heritage institutions. Source country capacity building — training local archaeologists and enforcement personnel, developing national site registries, and digitally documenting sites before they are at risk — is the highest-leverage investment the international community can make.

— ✦ —

What the past decade has established beyond reasonable doubt is that the archaeological record as it existed circa 2000 — dominated by excavated terrestrial sites in the temperate world, the Near East, and Mesoamerica — was not merely incomplete but constitutively misleading. It gave us a model of human prehistory built from the minority of the human past that happened to be accessible to scholars with trowels and conventional survey equipment. The emergence of LiDAR, SAR, and underwater survey as mainstream archaeological tools is not merely expanding our knowledge at the margins. It is replacing a partial map with an increasingly accurate one — and the shape of the new map looks very different from the old.

The civilizations were larger. The landscapes were more managed. The populations were greater. The coasts were inhabited far into what is now the sea. The forests hid cities. The deserts hid networks. And sub-Saharan Africa — the continent where our species originated and where it spent most of its evolutionary history — remains the most important blank on the archaeological map, not because nothing is there, but because we have barely looked.

The illusion of knowledge is dissolving. What replaces it will take a generation to fully understand.

■ Verified Sources & Formal Citations

1. Canuto, M.A., Estrada-Belli, F., Garrison, T.G., et al. (2018). "Ancient lowland Maya complexity as revealed by airborne laser scanning of northern Guatemala." Science, 361(6409). https://www.science.org/doi/10.1126/science.aau0137
2. Rostain, S., Dorison, A., De Saulieu, G., et al. (2024). "Two thousand years of garden urbanism in the Upper Amazon." Science, 383(6687), 183–189. https://www.science.org/content/article/laser-mapping-reveals-oldest-amazonian-cities-built-2500-years-ago
3. Prümers, H., Jaimes Betancourt, C., Iriarte, J., et al. (2022). "Lidar reveals pre-Hispanic low-density urbanism in the Bolivian Amazon." Nature, 606, 325–328. https://www.nature.com/articles/s41586-022-04780-4
4. Evans, D.H., Fletcher, R.J., Pottier, C., et al. (2013). "Uncovering archaeological landscapes at Angkor using lidar." Proceedings of the National Academy of Sciences, 110(31), 12595–12600. https://www.pnas.org/doi/10.1073/pnas.1306539110
5. Evans, D. (2016). "Airborne laser scanning as a method for exploring long-term socio-ecological dynamics in Cambodia." Journal of Archaeological Science, 74, 164–175. Summarized in: Smithsonian Magazine (2016): "Laser Scans Reveal Massive Khmer Cities Hidden in the Cambodian Jungle"
6. Estrada-Belli, F., Gilabert-Sansalvador, L., Canuto, M.A., et al. (2023). "Architecture, wealth and status in Classic Maya urbanism revealed by airborne lidar mapping." Journal of Archaeological Science, 157, 105835. Coverage: Tulane University News Release (2023)
7. Auld-Thomas, L. et al. (2024). Valeriana site formal identification. PACUNAM LiDAR Initiative ongoing reports. Coverage: Geo Week News (October 2025): "Spooky Signals: How Lidar Has Mapped Darker Sides of History"
8. Inomata, T. (2024). "Lidar, Space, and Time in Archaeology: Promises and Challenges." Annual Review of Anthropology, 53, 75–92. https://www.annualreviews.org/content/journals/10.1146/annurev-anthro-041222-093758
9. Mattingly, D.J. et al. (2011). "Lost Fortresses of Sahara Revealed by Satellites." National Geographic / Journal of African Archaeology. Coverage: National Geographic: "Lost Fortresses of Sahara Revealed by Satellites"
10. Gaffney, V. et al. (2025). "Final days of Doggerland interactive map and animation." University of Bradford, Submerged Landscapes Research Centre (Unpath'd Waters project). https://www.bradford.ac.uk/news/archive/2025/final-days-of-doggerland-interactive-map-and-animation.php
11. Gaffney, V. (2025). "Conjuring the Lost Land Beneath the North Sea." Hakai Magazine (with description of 2024 sediment core analysis). https://hakaimagazine.com/features/conjuring-the-lost-land-beneath-the-north-sea/
12. Bailey, G., Sakellariou, D., and SPLASHCOS Network (2012). "Submerged prehistoric archaeology and landscapes of the continental shelf." Antiquity Project Gallery, 86(334). https://antiquity.ac.uk/projgall/sakellariou334
13. Bailey, G.N., Flemming, N.C. (2008). "Archaeology of the continental shelf: marine resources, submerged landscapes and underwater archaeology." Quaternary Science Reviews, 27, 2153–2165. Overview in: Springer: "Archaeology of the Continental Shelf: Submerged Cultural Landscapes" (2022 update)
14. World Archaeological Congress 10 (WAC-10). (2024). Session proceedings: "Submerged Palaeolandscapes: The Underwater Archaeology of Deep Time." https://worldarchaeologicalcongress.com/wac10/submerged-palaeolandscapes-the-underwater-archaeology-of-deep-time/
15. Galili, E. et al. (multiple). Atlit-Yam excavation reports. Summary: Wikipedia "Atlit Yam" (citing primary Galili sources through 2023)
16. Davis, D.S. & Douglass, K. (2020). "Aerial and spaceborne remote sensing in African archaeology: a review." African Archaeological Review, 37(1), 9–24. Referenced in: Global Digital Heritage: "A Review of Satellite Remote Sensing Techniques" (2025)
17. Klehm, C. (2020). "Geospatial Analysis in African Archaeology: Current Theories, Topics, and Methods." African Archaeological Review. https://link.springer.com/article/10.1007/s10437-020-09376-9
18. MAEASaM Project (2024). Phase 1 Completion Report and Phase 2 Announcement. University of Cambridge, Department of Archaeology. https://www.arch.cam.ac.uk/research/projects/current-projects/mapping-africas-endangered-archaeological-sites-and-monuments
19. Casana, J. et al. (2015 / updated 2024). "Satellite Imagery-Based Analysis of Archaeological Looting in Syria." Near Eastern Archaeology, 78(3). https://www.journals.uchicago.edu/doi/10.5615/neareastarch.78.3.0142
20. UNESCO (2015 / updated 2024). "Cultural Heritage Objects: A Stake in Armed Conflicts." Official UNESCO release on illicit antiquities trafficking. https://www.unesco.org/en/articles/cultural-heritage-objects-stake-armed-conflicts
21. UN Security Council Resolution 2347 (2017). Protection of Cultural Heritage as Security Imperative. https://www.un.org/securitycouncil/content/s/res/2347-(2017)
22. Munawar, N.A. (2025). "A Decolonial Framework for Post-Conflict Syria." Journal of Social Archaeology (advance online). Analysis of UNSC resolution implementation consistency. https://www.tandfonline.com/doi/pdf/10.1080/00934690.2025.2605603
23. Isakhan, B. & González Zarandona, J.A. (2024). "Heritage Conflict and the Council: The UNSC, UNESCO, and the View from Iraq and Syria." International Journal of Cultural Property, 31(2). https://www.cambridge.org/core/journals/international-journal-of-cultural-property
24. Parcak, S.H. (2009; updated editions). Satellite Remote Sensing for Archaeology. London: Routledge. Background summary: NASA Earth Observatory: "Peering through the Sands of Time" (updated 2025)
25. Bernardini, F. et al. (2024). "LiDAR Applications in Archaeology: A Systematic Review." Archaeological Prospection, Wiley. https://onlinelibrary.wiley.com/doi/full/10.1002/arp.1931
26. Fouquet, M. et al. / SAMM (2024–25). Discovery of submerged megalithic structures off Sein Island, Brittany. International Journal of Nautical Archaeology. Coverage: 404 Media (December 2025): "Scientists Discover Massive Underwater Ruins That May Be a Lost City of Legend"
27. Rostain, S. et al. (2022). "Amazônia Revelada" project — Amazonia Revealed. National Geographic Society grant. Coverage: Jerusalem Post / JPost (February 2025): "Under the Jungles, Lidar Helps Researchers Find the Ancient Civilizations of the Amazon"
28. Zimpel Neto, C. and Neves, E. (2025). LiDAR discovery of lost Portuguese colony (Príncipe da Beira region), Amazon. Washington Post Investigative Report, April 10, 2025. https://www.washingtonpost.com/world/interactive/2025/brazil-amazon-discoveries-archaeology-lidar/
29. Estrada-Belli, F. et al. (2025). "New regional-scale Classic Maya population density estimates and settlement distribution models through airborne lidar scanning." Journal of Archaeological Science: Reports (July 2025). https://www.sciencedirect.com/science/article/pii/S2352409X25003219

© 2026 Archaeology Today  ·  All rights reserved  ·  Reproduction with attribution permitted  ·  ISSN 0278-2626 (print) / 1557-9484 (online)