New “eye-in-the-sky” chip technology easily tops radar
By John Edwards
July 31, 2012
With unmanned aerial vehicles (UAVs) taking to the skies in unprecedented numbers, aircraft survivability is a growing concern. “See and be seen” has long been a fundamental principal of air traffic control, but on pilotless aircraft some form of technology is essential for ensuring collision-free flying.
On many UAVs, radar is used to continuously check the skies for potential dangers. Yet this approach has a serious drawback that limits its effectiveness in today’s increasingly crowded airspaces. While radar excels at detecting large objects, such as mountains and larger aircraft, its resolution is too coarse to pick up many types of smaller threats, including birds and many types of UAVs.
A proposed solution
Researchers at Germany’s Fraunhofer Institute for Microelectronic Circuits and Systems feel they have developed a chip technology that is far more accurate than existing radar-based obstacle detections systems, yet is still small and light enough to be incorporated into many kinds of UAVs.
Fraunhofer IMS’s new complementary metal–oxide–semiconductor (CMOS) sensor can be incorporated into a special type of camera to create an entirely new anti-collision technology. The CMOS sensor can measure three-dimensional distances much more efficiently and with far more precision that conventional radar systems, says Werner Brockherde, Fraunhofer IMS’s head of development.
Like a sensor used in a digital camera that’s operating in a black-and-white setting, every pixel on the sensor is assigned a specific gray value. “[Additionally], each pixel is also assigned a distance value,” Brockherde says. “This enables [UAVs] to accurately determine their position in relation to other objects around them.”
The new technology promises major advantages over radar-based UAV anti-collision systems that calculate distances using reflected radio signals. “The sensor possesses a much higher local resolution,” says Brockherde. “Radar images would be far too coarse when operating in such near-field conditions.”
When installed inside a UAV that’s essentially performing as a flying robot, the new CMOS sensor-based technology is capable of identifying small objects measuring as little as 20 by 15 centimeters at ranges of up to 7.5 meters. This distance information is transmitted to the craft’s central processing unit at a rate of 12 images per second.
Even when there is light interference, such as when the UAV is flying directly into the sun, the sensor is capable of providing highly accurate images. The system incorporates a time-of-flight (TOF) process that uses light sources to emit short pulses that are reflected by objects and bounced back to the sensor. I
n order to prevent over-bright ambient light from masking the signal, the electronic shutter only opens for a few nanoseconds. For enhanced accuracy, the sensor also takes differential measurements—an initial image is captured using ambient light only; a second is then taken using the light pulse. The difference between the two measurement determines the final output signal. “All of this happens occurs in real time,” Brockherde observes.
The CMOS sensor research was conducted as part of the AVIGLE project, a winner of the Hightech.NRW technology competition, which receives funding from both the E.U. and the German state of North Rhine-Westphalia.
TriDiCam, a Fraunhofer IMS spinoff company, is already beginning to incorporate the CMOS sensor into a new line of cameras. System pricing and availability is expected to be announced shortly.