Daedalean
Daedalean, a Zurich company founded in 2016, builds machine-learning avionics that let aircraft navigate by camera, spot other traffic and assist with landing. Destinus Group completed its acquisition in January 2026.
Founded by former Google and SpaceX engineers, it has 51 to 200 staff. Its PilotEye system watches the full field of view for traffic that does not broadcast its position. Two EASA innovation partnership contracts in 2020 developed early concepts for certifying neural networks in aircraft.
Destinus paid 180 million Swiss francs in cash and shares and has made Daedalean the centre of its AI and autonomy work, with a shift towards defence. No airline or manufacturer customer is named, and a reported Honeywell partnership is unconfirmed.
No verified deployments yet
Daedalean is profiled in the companies register, but no deployment has been verified against published evidence. Records are added as evidence is reviewed and graded.
Products & Solutions
1 productPilotEye
AI-based visual traffic awareness and detection system jointly developed by Daedalean and Avidyne. Uses onboard cameras and a convolutional neural network to detect non-cooperative and cooperative airborne traffic (aircraft, drones, birds, gliders, parachutes) and alert pilots.
Pilots cannot continuously monitor the full field of view for non-cooperative traffic (aircraft without ADS-B, drones, birds, gliders, parachutes). Existing traffic advisory systems rely on transponder or ADS-B signals and miss non-equipped aircraft. PilotEye provides a dissimilar, redundant detection layer that does not depend on other aircraft broadcasting their position.
- Visual detection of non-cooperative traffic including aircraft without ADS-B, drones, birds, gliders, and parachutes
- Detection range up to 3.5 km
- Up to 10 Hz update rate
- Traffic classification, identifies size, position, direction of movement, and time to collision
7 more features
- Integration with Avidyne IFD and SkyTrax avionics for cockpit display
- Capstone-compliant target information and thumbnail video output
- Deterministic (machine-learned, not machine-learning), neural networks fixed after training, no in-flight learning
- Wire detection capability (detecting wires only a few pixels wide)
- All computation performed onboard, no ground infrastructure dependency
- Emergency landing zone guidance and runway/landing zone guidance
- Ground taxi guidance and traffic detection
Technical specifications (10)
- Hardware: 1x VXP (Visual Everything Processor) + 3x VCU (Visual Capturing Unit)
- Processor weight: 8.50 lbs (3.85 kg); Camera weight: 0.80 lbs (0.36 kg) per unit
- Processor dimensions: 7.04 x 2.96 x 12.84 inches (17.88 x 7.51 x 32.61 cm)
- Camera dimensions: 2.66 x 3.25 x 6.00 inches (6.75 x 8.25 x 15.24 cm)
- Power: 11-33 VDC, 2.2A at 28 VDC
- Environmental: DO-160G qualified, to 50,000 ft, -20C to +55C operating, -40C to +70C short-term
- No external cooling required
- Software: DO-178C DAL-C (DAL-B planned)
- Hardware: DO-254 DAL-C (DAL-B planned)
- Architecture: high-resolution avionics-grade cameras with IMUs, convolutional neural network core, classical pre/post-processing software
- STC application in progress, concurrent FAA and EASA certification (SOI stages 1 and 2 of 4 completed as of October 2025; SOI stage 3 commenced summer 2025)
- FAA classification: machine-learned (not machine-learning), deterministic behavior
- EASA/FAA joint CRI (Certification Review Item) and IP (Issue Paper) issued
- Daedalean certification methodology referenced in EASA 'First Usable Guidance for Level 1 AI/ML applications' (CoDANN framework)
Profile data from the Aerostack companies register; deployment records from the verified register. Evidence grades: T1 first-party confirmation, T2 credible reporting. See Our approach for how records are graded.