Lockheed Martin's Skunk Works division has completed a landmark test series in which an autonomous fighter aircraft executed 27 live intercepts relying solely on a passive infrared sensor, without radar emissions or synthetic training data. The experiment, designated Have Heat, establishes the first flight-tested baseline for artificial intelligence combat operations in electronically contested environments where adversary jamming would render traditional active sensors ineffective. The demonstration marks a significant step toward operationalizing AI-driven tactical aviation in scenarios that mimic the electronic warfare capabilities fielded by China and Russia.
The test campaign utilized the X-62A VISTA (Variable In-flight Simulator Test Aircraft), a modified F-16 platform that serves as the U.S. Air Force's primary testbed for autonomous flight algorithms. During the Have Heat series, the AI agent processed only passive infrared signatures to detect, track, and engage target aircraft across 27 separate intercept scenarios. No radar transmissions were made by the test aircraft, and the infrared sensor operated without emitting any detectable energy signature, a critical attribute for survival in dense jamming environments where active sensors immediately reveal platform position.
What's New / Specs
The Have Heat test series represents several firsts in the integration of artificial intelligence with tactical fighter aircraft. The core innovation lies in the exclusive reliance on passive infrared sensing for the complete kill chain — detection, tracking, weapons employment guidance, and post-intercept assessment — without any radar cueing or datalink support. This approach directly addresses a persistent vulnerability in modern air combat: active electronically scanned array (AESA) radars, while highly capable, emit energy that can be detected, geolocated, and targeted by adversary electronic support measures (ESM) and anti-radiation missiles.
- Platform: X-62A VISTA (Variable In-flight Simulator Test Aircraft), a highly modified F-16D operated by the U.S. Air Force Test Pilot School at Edwards Air Force Base
- Sensor suite: Passive infrared search and track (IRST) system, specific model not disclosed; no radar emissions during intercepts
- Test designation: Have Heat, conducted by Lockheed Martin Skunk Works in collaboration with the Air Force Research Laboratory and Air Force Test Center
- Intercept count: 27 live intercepts completed across the test series
- AI autonomy level: Full autonomous detection, tracking, and weapons employment guidance; human safety pilot onboard with override authority
- Environment: Electronically contested, simulating peer-adversary jamming that would deny radar and datalink use
- Data sources: Live sensor data only; no synthetic training data used during the test execution phase
The X-62A VISTA platform has been central to the Air Force's autonomy development since its redesignation from the NF-16D in 2021. The aircraft's fly-by-wire system and variable stability architecture allow researchers to upload and test autonomous control laws in a safe, controlled manner while retaining a qualified test pilot in the cockpit for safety oversight. Previous VISTA campaigns, including the February 2026 demonstration reported by the American Institute of Aeronautics and Astronautics (AIAA), have shown the platform evading simulated incoming missiles using onboard tactical AI. The Have Heat series extends this lineage by focusing specifically on the sensor-to-shooter chain under electronic attack.
Why It Matters
The strategic significance of the Have Heat demonstration extends well beyond a single test series. Peer competitors, particularly China and Russia, have invested heavily in electronic warfare capabilities designed to deny, degrade, or deceive the active sensors and communication networks that underpin current U.S. and allied air superiority concepts. The People's Liberation Army Air Force fields advanced jamming platforms such as the J-16D and dedicated electronic attack variants of the J-20, while Russian Su-35S and Su-57 platforms integrate powerful digital radio frequency memory (DRFM) jammers capable of saturating X-band and Ku-band radar frequencies. In a high-end conflict, the assumption of unfettered radar use and reliable datalinks is increasingly untenable.
Passive infrared sensing offers a compelling countermeasure. Because IRST systems detect thermal emissions rather than transmitting radio frequency energy, they are inherently immune to radar jamming, anti-radiation missiles, and passive detection by adversary ESM. However, infrared sensors historically suffer from shorter range, weather sensitivity, and the absence of precise range measurement without laser ranging or triangulation. The Have Heat results suggest that modern AI processing — likely employing advanced convolutional neural networks for target detection and recursive Bayesian filters for track maintenance — can overcome these limitations sufficiently to support weapons-quality tracks in realistic engagement geometries.
The implications for force design are substantial. If autonomous fighters can reliably execute intercepts using only passive sensors, the Air Force gains options for distributed, low-observable operations in denied environments. Collaborative combat aircraft (CCA) concepts, which envision loyal wingman drones operating alongside crewed fighters, could leverage similar passive sensor suites to reduce their electromagnetic signature and improve survivability. The Have Heat baseline also informs requirements for next-generation air dominance platforms, where sensor fusion across passive RF, infrared, and electro-optical modalities will be essential.
Our Take
The Have Heat demonstration is a credible, measurable advance in the maturation of autonomous air combat. By constraining the test to 27 live intercepts with a single passive sensor modality, Skunk Works and the Air Force have produced a clean, verifiable data point that avoids the ambiguity of simulation-heavy claims. The exclusion of synthetic data during execution is particularly noteworthy; it forces the AI to contend with real-world sensor noise, atmospheric effects, and target maneuvers that synthetic datasets often smooth over.
However, several caveats temper enthusiasm. First, the test environment, while electronically contested, likely did not replicate the full density and sophistication of a peer adversary's integrated air defense system, which would combine jamming with decoys, cyber effects, and kinetic threats. Second, passive infrared range performance against low-signature targets — such as stealthy cruise missiles or fifth-generation fighters with reduced thermal signatures — remains an open question. Third, the transition from a testbed like VISTA to an operational CCA or next-generation fighter involves certification, cybersecurity, and logistics hurdles that are non-trivial. The Air Force's own autonomy roadmap acknowledges that software certification for safety-critical AI remains a pacing challenge.
Editorially, the Have Heat series should be viewed as a necessary and successful proof-of-concept, not an operational capability. The next logical steps — multi-ship autonomous coordination under jamming, integration with offboard sensors via low-probability-of-intercept datalinks, and demonstration against representative threat replicas — will determine whether this approach scales to warfighting relevance. For now, the 27 intercepts stand as the most concrete evidence yet that AI can close the sensor-to-shooter loop without radiating.
FAQ
What aircraft was used for the Have Heat test series?
The X-62A VISTA (Variable In-flight Simulator Test Aircraft), a highly modified F-16D operated by the U.S. Air Force Test Pilot School at Edwards Air Force Base, served as the test platform. The VISTA's variable stability system allows researchers to test autonomous control laws while a safety pilot retains override authority.
How many intercepts were completed and what sensor was used?
The test series completed 27 live intercepts using only a passive infrared search and track (IRST) sensor. No radar emissions were made by the test aircraft during any intercept, and no synthetic training data was used during test execution.
Why is passive infrared sensing important for future air combat?
Passive infrared sensors detect thermal emissions without transmitting radio frequency energy, making them immune to radar jamming, anti-radiation missiles, and detection by adversary electronic support measures. This capability is critical in electronically contested environments where peer adversaries like China and Russia field advanced jamming platforms.
Does this mean AI fighters are ready for combat deployment?
No. The Have Heat demonstration is a proof-of-concept establishing a flight-tested baseline. Operational deployment requires further development in multi-ship coordination, integration with offboard sensors, certification of safety-critical AI software, and testing against representative threat systems in denser electronic warfare environments.
How does this relate to the Air Force's Collaborative Combat Aircraft program?
The Have Heat results inform CCA concepts by demonstrating that autonomous aircraft can execute intercepts using only passive sensors, reducing their electromagnetic signature and improving survivability. Future loyal wingman drones could employ similar sensor suites to operate effectively in denied environments alongside crewed fighters.
