Since 2016, Daedalean has been developing a suite of pilot assistance and autonomous flight applications by fusing Artificial Intelligence with advanced electro-optical sensor technologies. In parallel, the company has worked closely with regulators to lay the groundwork for the safety assurance of AI-based technologies. Now, Daedalean is on track to make history later this year as the first to certify a safety-critical system featuring an AI-enabled component.
Artificial intelligence has ignited productivity and efficiency across various sectors, and Daedalean’s mission is to harness that computing power to enhance safety in aviation. The company began by using AI-powered computer vision to perform tasks a pilot carries out in visual flight conditions. Aircraft-mounted cameras feed an onboard, AI-enhanced computer to create real-time, comprehensive situational awareness. The next stage will be fusing information from more sensors (such as night vision cameras and radar) to create awareness in all conditions. And eventually, the capabilities will go beyond awareness—Daedalean’s systems will be able to anticipate and react to future situations, including potential threats. The company refers to this capability as Situational Intelligence.
Real-World Trials
Daedalean took a big leap forward this year by flight testing, with Moog, one of its most advanced systems. As announced last year, Moog and Daedalean embarked on a joint project to flight test and evaluate Daedalean’s visual awareness system for integration into Moog’s Genesys EFIS display product line.
Recently, the two companies took to the air, as Daedalean’s system interfaced with the Genesys IDU-680 flight management system, making possible closed-loop, GPS-free autonomous navigation for a Bell OH58 test platform. Navigation was provided by Daedalean’s system, locking onto geographic features on the ground to determine position instead of relying on satellite-based information. In a world where GPS can be easily spoofed or jammed, this capability has obvious and widespread applications.
The flight testing also featured two other key Daedalean functions: traffic detection and landing guidance.
In mission scenarios where a landing site cannot be predetermined, the system autonomously identifies suitable landing zones in real time. (For current testing, sites are indicated by an AprilTag, similar to a QR code, but future versions will identify remote landing zones automatically.) Once a viable site is detected, it is automatically uploaded as a waypoint into the Flight Management System, which then dynamically generates a new flight path and executes a precision landing. Unlike standard GPS-based approaches—which can have horizontal errors exceeding ±5 meters—this system enables far more accurate landings, a necessity when attempting to land unmanned rotary-wing aircraft in constrained environments.
Daedalean’s traffic detection technology is among the company’s most mature, enabling identification of aerial objects, including non-cooperative traffic (such as drones, birds, balloons, paragliders, etc.) without the need for ADS-B, ATC, or radar. As a pilot assistance application, it can significantly increase safety for General Aviation aircraft. Such a system will provide a pilot with a reliable and permanent flow of information, seeing farther and better than a pilot. The AI can identify an object from just a few pixels in a single frame—far less detail than the human eye can distinguish. This allows it to detect potential hazards early, such as a distant aircraft appearing as a tiny dot but actually on a collision course or an object blending into a dark landscape below the horizon. But before it can assist pilots in the cockpit, the technology must first be certified.
Certifying Trustworthy AI
To date, no AI-based safety-critical systems have been certified for civil aviation. Regulators are actively seeking frameworks to ensure such systems would be safe, reliable, and fit for purpose. Traditional software compliance methods fall short for machine-learned systems. As the FAA notes, “the implementation is learned, rather than designed,” meaning the code cannot be easily traced back to explicit requirements or validated through conventional means. However, both the FAA and EASA are advancing in the adoption of new technologies. In 2024, EASA took a major step by expanding its guidance with the release of "AI Concept Paper Issue 2." Meanwhile, the FAA published its "Roadmap for AI Safety Assurance." These efforts signal growing alignment between regulators and industry on how to integrate AI safety assurance into the traditional design assurance process.
Daedalean is well on its way to certifying the first real-world application of a system with an AI component. The company is working with the FAA and EASA to certify two versions of its visual traffic detection capability: PilotEye for fixed-wing and Ailumina Vista for rotary-wing aircraft. Ailumina refers to Daedalean’s entire line of situational awareness products, while PilotEye originated as a jointly developed product with the US-based avionics manufacturer Avidyne. Both systems are to be certified to DAL-C and benefit from concurrent validation by the counterpart regulatory agency. Delegations from both regulators have already conducted several week-long visits to Daedalean’s offices, completing two of four Stage of Involvement audits. Certification is expected to be attained by the end of 2025.
Visual Technologies Advancing Safety
In parallel, Daedalean is advancing its system from simple detection to full detect-and-avoid capability by adding avoidance maneuver guidance. This is a critical enabler for beyond visual line of sight (BVLOS) operations—one of the major hurdles to integrating uncrewed aircraft into the National Airspace System. With this capability, remotely piloted aircraft—from helicopters and cargo planes to high-payload drones—can begin to operate autonomously in missions like offshore transport, aerial ambulance, search and rescue, and cargo delivery, making such operations safer, more cost-effective, and more scalable.
Daedalean’s visual traffic detection technology also powers a drone protection system. This passive, AI-driven threat detection module—integrated with EO/IR cameras—can be mounted on fixed infrastructure (like towers, rooftops, or poles) or deployed on patrolling aircraft and drones. It can continuously monitor the airspace around high-risk facilities, providing 24/7 operational vigilance and early warning of hostile drone activity. Because it relies solely on vision—no radar, no emissions—the system is invisible to intruder detection and scanning—creating a protective “safety dome” without revealing its presence.
And AI's ability to interpret visual information opens the door to a wide range of safety-critical capabilities—one of the most impactful being obstacle detection.
Power lines, including both wires and masts, pose a persistent threat to low-flying aircraft, with around 30% of collisions proving fatal. Existing detection methods are limited. Some rely on static databases that can’t account for real-time changes, while others depend on electromagnetic (EM) signals, rendering them useless for unpowered wires.
Visually detecting these hazards is especially challenging, as wires often lack distinguishing features and can’t be tracked as single, identifiable points. Daedalean has developed a solution that overcomes these limitations with an AI-powered system capable of identifying masts and wires even when they are effectively invisible to the human eye. It works independently of external databases or EM field emissions. Still in development, the system has already exceeded expectations in test flights, demonstrating detection ranges of several hundred meters and impressive accuracy, even when key visual features span just a few pixels in a single frame of video.
The safety impact is substantial: this technology gives pilots real-time alerts of life-threatening obstacles that would otherwise go unnoticed—especially in low-altitude or off-airport operations.
Object detection isn’t limited to static hazards. AI-powered vision systems can dynamically identify moving threats as well—an especially critical capability during landing.
One of the most pressing safety challenges in aviation is runway incursion. And when an aircraft lands on an uncontrolled runway, the risk of collision or near-miss with another aircraft, a vehicle, an animal, or even a person increases dramatically.
This risk isn't limited to general aviation. Business jets frequently land at smaller airports that lack control towers or Instrument Landing Systems (ILS)—infrastructure typically reserved for large commercial hubs. For perspective, in the United States, fewer than 10% of the 5,000+ public airports have control towers, and there are nearly 20,000 private airfields, most of them uncontrolled.
Daedalean is developing a vision-based AI-enabled system that can detect and alert pilots to the presence of any object—moving or stationary—on or near the runway. This could dramatically enhance situational awareness and landing safety across all aircraft categories.
Daedalean’s AI technology is delivering real, immediate safety enhancements—like traffic detection, GPS-free navigation, and landing hazard alerts—that can be deployed in today’s cockpits. At the same time, as certification approaches, it’s paving the regulatory and technical path for AI to power the next wave of aviation innovation, unlocking new capabilities for the entire industry.