Xwing, Daedalean Collaborate “to Advance AI in Aviation” 

Xwing, a sup­pli­er of mod­u­lar auton­o­my tech­nol­o­gy for avi­a­tion, and Daedalean AG, a devel­op­er of safe­ty-crit­i­cal and cer­ti­fi­able Arti­fi­cial Intel­li­gence (AI) sys­tems for sit­u­a­tion­al aware­ness and flight con­trol, recent­ly announced a for­mal col­lab­o­ra­tion to share data, knowl­edge and process­es in the field of machine learn­ing and AI, to accel­er­ate time-to-mar­ket while enhanc­ing safe­ty, reports a press release. The part­ners are work­ing close­ly with the FAA and EASA to cer­ti­fy their soft­ware.

The release explains, “No machine learn­ing (ML) based safe­ty-crit­i­cal appli­ca­tions have made their way into the cock­pit of civ­il avi­a­tion so far – this requires a par­a­digm shift in how soft­ware and hard­ware have tra­di­tion­al­ly been assessed and cer­ti­fied as safe.”

It con­tin­ues, “Avi­a­tion reg­u­la­tors have been work­ing on how machine learn­ing fits into their design assur­ance process­es and how to adjust the respec­tive stan­dards. Both com­pa­nies believe that form­ing a con­sen­sus on their assur­ance approach­es and shar­ing infor­ma­tion is the best way to advance the devel­op­ment of cer­ti­fi­ca­tion stan­dards that will allow an increase in safe­ty more effec­tive­ly than if they worked sep­a­rate­ly.” 

Luuk van Dijk, co-Founder and CEO of Daedalean, com­ment­ed, “With this shared under­tak­ing, we will be able to demon­strate that increas­ing safe­ty is dri­ving inno­va­tion and that a col­lab­o­ra­tive approach to har­monise reg­u­la­tions and stan­dards ensures that best prac­tices are uni­ver­sal­ly adopt­ed.” 

Xwing unveils autonomous flight sys­tem for region­al planes (Cred­it: Xwing)

Both Xwing and Daedalean have pub­lished award-win­ning papers that out­line their approach­es to cer­ti­fi­ca­tion. For exam­ple, Xwing’s For­mal and Prac­ti­cal Ele­ments for the Cer­ti­fi­ca­tion of Machine Learn­ing Sys­tems as well as Daedalean’s Con­cepts of Design Assur­ance for Neur­al Net­works series. The col­lab­o­ra­tion will help dri­ve progress towards the estab­lish­ment of an accept­able means of com­pli­ance. 

Maxime Gariel, Pres­i­dent and CTO of Xwing, remarked, “The col­lab­o­ra­tion under­scores a phi­los­o­phy of shar­ing data, knowl­edge and process­es toward cer­ti­fi­ca­tion for the indus­try as a whole as we work close­ly with reg­u­la­tors.” 

The release con­tin­ues, “While Daedalean and Xwing have inde­pen­dent­ly devel­oped frame­works for the cer­ti­fi­ca­tion of AI-based sys­tems, both com­pa­nies agree that high­er lev­els of safe­ty can be achieved on a short­er time­line by shar­ing knowl­edge and rel­e­vant flight data, so that all stake­hold­ers, both indus­try and reg­u­la­tors as well as the pub­lic, can make sig­nif­i­cant progress on the path to cer­ti­fy­ing AI sys­tems. 

Daedalean AI for Flight Con­trol and Sit­u­a­tion­al Aware­ness (Cred­it: Daedalean)


Found­ed in 2016, Xwing is based in North­ern Cal­i­for­nia with offices in San Fran­cis­co and Con­cord.  It is an indus­try’s sup­pli­er of mod­u­lar auton­o­my tech­nol­o­gy for avi­a­tion with adapt­able sys­tems designed to inte­grate with a wide range of air­craft, pro­vid­ing the abil­i­ty to safe­ly oper­ate ground-super­vised, uncrewed air­craft in any air­space.

Daedalean was found­ed in 2016 and has offices in both the EU and U.S. The team of over 150 peo­ple, includ­ing 15 PhDs and 19 pilots, pos­sess exper­tise in machine learn­ing, avi­a­tion-grade soft­ware and com­put­ing hard­ware engi­neer­ing, flight test­ing, safe­ty assess­ment, and cer­ti­fi­ca­tion.

The com­pa­ny in col­lab­o­ra­tion with Avi­dyne, is work­ing to cer­ti­fy and bring to the mar­ket the first cer­ti­fied prod­uct for Gen­er­al Avi­a­tion: Pilot Eye­TM, a visu­al traf­fic detec­tion source capa­ble of detect­ing non-coop­er­a­tive traf­fic.

For more infor­ma­tion



(Top image: Cred­it — Xwing/Daedalean)

eVTOL Insights is part of the Industry Insights Group. Registered in the UK. Company No: 14395769