After more than a century of powered aviation the navigation, guidance and control of aircraft is still a very manual affair, with many functions relying directly on the human pilot. This presents a technical hurdle to expanding the economic applications of flight. It is bound to become a more pressing problem with more than 100 companies trying to bring small (mostly electric) VTOL aircraft ("flying cars") through certification and to market. Replacing the human in the loop with modern robotic systems will become a crucial enabler for whole new markets in aviation and urban transport.
Safety is tightly regulated in aviation and rightly so because it has made flying one of the safer means of transportation. Yet these regulations and the built-in desire for proven solutions also present a barrier to the adoption of new technology. There is no incremental path from uncertified UAV control to certification for personal transport (DAL-A), nor one from current DAL-A level avionics to full autonomy. To justify the absence of a human pilot we need to build a system that is sufficiently deterministic to pass certification, yet can deal with unexpected situations.
Relying on our expertise in robotics, computer vision and machine learning, as well as a thorough foundation in classical systems engineering, avionics and piloting, we have set out to build guidance navigation and control systems that replace and outperform the human pilot on every measurable scale. To autonomously fly in VFR, without relying on extra rules or infrastructure, requires that navigation, guidance and collision avoidance are able to use visual information first and last, alongside other instruments currently available in the cockpit. Our visual systems provide situational awareness and semantic understanding to safely guide any "flying car" from take-off to landing.