We began as a research project almost a decade ago at Ben-Gurion University’s Electrical & Computer Engineering Department, exploring the area of real-time registration. In the course of the research, we realized the incredible potential it had for helping to make autonomous flight a reality. Our fundamental challenge is this: how to analyze a drone’s ever-changing surroundings accurately and in real-time while using a minimum of hardware, computer power and network bandwidth. On top of that, take-offs and landings have to be reliably safe, and a truly autonomous drone has to perform its functions without relying on ground-based resources.
Our patented innovation was to use real-time registration to give flying vehicles the ability to see where they are and understand their surroundings, by comparing what they see in real time to a reference frame or a map to successfully navigate to their destination and analyze a given scene with the highest accuracy possible. After years of research and testing, we delivered our first real-time autonomous flying system to Israeli Aerospace Industries (IAI), a major player in aviation. This success was the first step in showing that our innovative research had powerful real-world applications for autonomous flights.
Today, we’re working with a number of autonomous drone systems to extend where drones can go and what they are capable of. The key to it all is something called situational awareness: our technologies enable drones and flying vehicles to understand where they are based on what they see, at a range of altitudes and with higher precision than that of a GPS, or without GPS at all. Achieving this is a significant challenge, and the underlying technology it requires is correspondingly impressive. We are constantly working on improving image- and point cloud-matching algorithms, to make them more powerful and compact. But we’re also exploring how other AI technologies can improve air mobility, too. Like how we can use big data to make drones fly with greater precision, and how to use deep learning to teach drones how to understand their surroundings more precisely, all in real time.
At Sightec, we’re practical dreamers. Scientific visionaries. Engineers and pioneers. We’re fearlessly optimistic – both of those words are essential. Because of course it takes extremely complex mathematics and algorithms to achieve what we’re trying to achieve, but it takes something else, too. It takes an unflinching belief in the idea of human progress – a commitment to harnessing the most sophisticated technology available for the purpose of making everyday life a little bit better.