Software developer Daedalean is making inroads on development of artificial intelligence (AI) tools that power pilot assistance and eventually autonomous flight operations. These tools will be incorporated into upcoming avionics products and aircraft.
Daedalean uses the term “situational intelligence” to describe how its AI tools will help pilots. What Daedalean is not doing is creating “this massive AI brain, and out comes this magical situational intelligence,” said Luuk van Dijk, founder and CEO of the Switzerland-based company. “We have to work within the bounds of certifiability. We’ll use it for functions that are hard to capture in traditional software.”
AI is needed because traditional software doesn’t have the capability, for example, to look at photos of airplanes and distinguish which shows an airplane and whether the airplane is on a runway. This skill is key to the work that Daedalean is advancing to use AI to spot aircraft and help them navigate safely around other traffic. “That’s hard to capture without AI techniques,” van Dijk said.
“The product roadmap initiates with pilot assistance, then single-piloted vehicles, then full autonomy,” said Yemaya Bordain, president of Daedalean’s Americas business. One product will be an AI-enhanced flight planning system that integrates with the aircraft’s autopilot and enables autonomous navigation and dynamic updates in real time. This will ensure an aircraft can safely travel through designated corridors like, for example, an eVTOL flying at low altitudes.
“We’re simultaneously developing enabling technologies to facilitate autonomy,” she said, and this is being done in collaboration with regulators such as the FAA and EASA.
On the hardware side, Daedalean is designing high-performance computers and sensors such as high-resolution cameras and radar. “We’re also working on situational intelligence," Bordain said, "applying AI in applications that today and in recent years were key enablement capabilities that have been reserved only to humans.”
One of the early Daedalean products will be PilotEye, a machine-learning traffic awareness system, which is being developed in partnership with avionics manufacturer Avidyne. “We’re targeting to have this certified by the end of this year or early next year,” she said.
Bordain said that Daedalean has applied to become an EASA design organization. “We expect we will be the only company in the world that knows how to certify machine learning by earning that designation.”
Another avionics manufacturer that has partnered with Daedalean is Moog’s Genesys Aerosystems unit. “Their go-to-market aligns well with what we want to do,” she said. “We see ourselves as a company that can work across the value chain, with avionics makers, airframers, and through partners to provide whole systems.”
“These are not R&D projects, but products,” van Dijk said. “Partnering with Moog and Genesys, it’s nice to have a brain [from our system] but we need to have an autopilot and FMS to integrate with. We can immediately address all platforms they put their FMS and autopilot in.”