Developing Automated Conflict Avoidance for Drones
Carlyn McGill posted on December 16, 2015 |
Unmanned aircraft flying above trees.
Unmanned aircraft flying above trees.
Stanford researchers have developed new software to predict and prevent unmanned aircraft collisions when drones are flying at low altitudes.

The unnamed software isolates issues involving multiple drones and formulates alternative flight paths for selected pairs. A server then coordinates each of these solutions and gives a joint collision avoidance order.

To establish just how many drones could be contacted at once, researchers ran more than one million simulations where between two and ten unmanned aircraft encountered each other while in flight, risking collision. The paired strategy was compared to other strategies, including one where the drone only reacts to its closest threat. The solution with paired drones proved to be safer, and offered faster decision times and decreased alert rates.

The new software can predict and prevent drone collisions.

The new software can predict and prevent drone collisions.

The Stanford Intelligent Systems Laboratory (SISL) is working with the NASA Ames to develop this experimental Unmanned Aerial System Traffic Management system (UTM). The system is being developed to maintain and control traffic with the expected future increase in low altitude unmanned flights.

The UTM system is projected to take over many of the responsibilities and functions currently performed by air traffic control. Requiring air traffic control operators monitoring all the actions of drone operations would be unfeasible.

Stanford University calculates that the Federal Aviation Administration has approximately 15,000 human controllers who manage about 87,000 pilot-driven flights daily. When these numbers are compared to what it would take to monitor potentially hundreds of thousands of unmanned aircraft flights a day, it’s simply impractical.

Collisions between more than two aircraft are not very common, but flights in small and crowded urban areas create an opportunity for multiple aircrafts to have a collision. Researchers believe that automated conflict avoidance is the best solution to regulate traffic and prevent collisions.

The new software hopes to ensure safer flight for drones in crowded urban areas.

The new software hopes to ensure safer flight for drones in crowded urban areas.

There is still work left to be done. Communication breakdowns, bad weather and disruptive drones are problems that will still need to be addressed.

NASA estimates the final builds of the UTM will be completed by 2019. For more information on the UTM system, visit their website.


Recommended For You