Sense and Avoid Sensor Selection
Developing sense-and-avoid (SAA) capabilities are challenging for small unmanned aircraft. Typical onboard systems for detecting intruding aircraft include non-cooperative systems such as radar, or cooperative systems like Traffic Collision Avoidance System (TCAS). These options are currently not suitable for sUAS due to relatively large size, weight, power consumption, and cost. Benefits of the sUAS SAA sensor trade-space include short ranges and low speeds, which enable vision-based systems of low-cost electro-optical (EO) or infrared (IR) sensors to be used.
Sensor Recommendation
The proposed sensor system would be based on research into the optical flow sensed by insects and used for navigation (Barnhart, R. K., Shappee, E., & Marshall, D. M., 2011). In an example case, a house fly uses thousands of simple receptor cells that detect changes in contrast. Those signals are sent to a section of the brain that contains elementary motion detectors, which build a relative motion map within the fly’s field of view. This system was recreated in a lab setup by Ruffier and Franceschini in 2008 using a tethered remote-controlled helicopter (referred to as OCTAVE2) testbed with the intent of avoiding terrain and close-in obstacles. The technique could also be adapted to provide an SAA capability between non-cooperating vehicles operating in the sUAS realm (low-speed, less than 400ft above ground level). The basic sensing element consisted of a pair of P-type, intrinsic semiconductor, N-type (PIN) diodes which conduct current upon reaching a certain threshold of received light (First Sensor AG, 2015). The PIN diodes were installed behind a 5mm diameter lens and mounted to a 400um-thick circuit board. The complete sensing element measures 2.7x3.0cm and weighs 4.3g. While a power requirement for this particular arrangement was not provided for the OCTAVE2, PIN diodes are available in a variety of supply voltages that would be compatible with sUAS (3.3-9V). Signals were routed to a microprocessor for running optical flow analysis, resulting in command generation for the autopilot. The system was found suitable for directing the OCTAVE2 over steep terrain to a safe landing with only two “eyes,” and could be further enhanced by with a higher quantity of smaller sensing elements, with the net effect of increasing resolution and field of regard. Depending on the mission of the sUAS, the SAA systems could also a subsystem of the overall vehicle navigation systems, instead of an addition, further reducing size, weight, and power requirements. This SAA technique has the strong advantages of being light weight, small, and requires little power and no additional sensing input (i.e. GPS, INS, pitot-static) to guide the vehicle. One disadvantage is relatively short range, which reduces the time available to alter the flight path. This will likely not be a problem for highly maneuverable quad-copters, but may result in the insect-inspired system being less desirable for fixed-wing sUAS.
References
Barnhart, R. K., Shappee, E., & Marshall, D. M. (2011). Introduction to Unmanned Aircraft Systems. London, GBR: CRC Press. Retrieved from http://www.ebrary.com
First Sensor AG. (2015). PIN photodiodes. Retrieved from http://www.first-sensor.com/en/products/optical-sensors/detectors/pin-photodiodes/
Ruffier, F., & Franceschini, N. (2008). Aerial robot piloted in steep relief by optic flow sensors. Paper presented at the 1266-1273. doi:10.1109/IROS.2008.4651089
