Sense and Avoid Collision for sUAS

Eli Hini
Eli Hini
Jul 12, 2018 · 3 min read

Introduction

One of the critical requirements for the successful integration of unmanned aerial systems into civilian airspace to operate side-by-side with other airborne traffic is complete sense and avoid capabilities otherwise known as collision avoidance (Jacob et. al, 2018). There are several works underway to bring this to bear. The integration of ADS-B (Automatic Dependent Surveillance — Broadcast) transponders is gaining attention. ADS-B allows aircrafts to broadcast their current state so that nearby aircrafts and air-traffic control systems can pick up on the communication to determine the location and heading of the broadcasting aircraft (FAA, n.d). Radio based solutions like ADS-B and radar are a good solution towards air traffic monitoring, however they do not provide visual confirmation to operators (Jacob et. al, 2018). As the airspace becomes busy, there is a need for other solutions that will directly benefit each aircraft in regard to seeing airborne obstacles within immediate vicinity and trajectory of travel. For example, being able to see a bird or other unmanned aerial vehicles (UAVs) and avoid them is critical to keeping the airspace safe. Iris Automation, has developed a collision avoidance system that relies on vision. Using computer vision and deep learning, Iris Automation is able to detect, classify and avoid airborne obstacles.

Iris Collision Avoidance System

Recent advancement in the development of computer vision and deep learning system has now made it possible to process scenes and objects in images and videos to produce actionable intelligence. Iris automation, based in Vancouver, British Columbia, leverages this advancement with their own proprietary algorithms to develop the Iris Collision Avoidance System (ICAS). ICAS is made up of a set of cameras, processing unit and software that can be installed on a UAV to enable it to see its environment. For rea-time results, the processing of captured data is done onboard the system.

The system can detect objects in and around the path of travel. One of its key features is its ability to detect the direction, distance and estimated speed of travel of other airborne obstacles at a range of 500m from the observing UAV. Upon detection, it can estimate a collision course and suggest alternate routes to avoid the collision. Relying on deep learning, the system can not only detect that there’s an intruding airborne obstacle but it can also identify them. It can identify and classify some planes and birds. This is very important for increasing the situational awareness of ground control crews.

The Iris Collision Avoidance System features a modular and highly configurable design. This makes it easy for other UAV manufactures to be able to adopt and integrate it into their UAVs. It also enables third party and aftermarket installation onto UAVs that don’t have them built-in by the manufacturers. ICAS can be integrated onto a system in three primary ways: (i) self-contained plug and play (ii) bare board for higher customization into any design and (iii) software only for systems that meet deep requirements (Iris Automation, n.d). ICAS can be extended further through its open application programming interface (API).

Summary

ICAS is an ideal solution for both existing and future small UAS (sUAS) because of its modular design and flexible integration option. OEMs and operators can select an integration option that meet their needs. As a full sense and avoid system, ICAS can be added as part of any suite of UAV sensors to complement safe flight operations without needing to task primary sensor payloads. With its open API, ICAS can be extended to fuse with ADS-B and other sense and avoid solutions to create an even safer and comprehensive collision and avoidance systems for both small unmanned aerial systems.

References

FAA. (n.d). Frequently Asked Questions. Retrieved from https://www.faa.gov/nextgen/programs/adsb/faq/#g1

Jacob, J., Mitchell, T., Loffi, J., Vance, M., & Wallace, R. (2018). Airborne visual detection of small unmanned aircraft systems with and without ADS-B. Paper presented at the 749–756. doi:10.1109/PLANS.2018.8373450

Iris Automation. (n.d). See Everything. Retrieved from https://www.irisonboard.com/technology/

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch

Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore

Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade