This is a summary of exaQuark, a technology for self-organized UAS traffic management our submission to the Ministry of Transport and Civil Aviation Authority of Singapore CFP
Unmanned Traffic Management
Collaborative collision avoidance, swarming and other decentralized collective behaviors rely on proximity communications: when drones are nearby they talk one to another to avoid collisions, fly together in an orderly manner or team up on a task.
Today, proximity communications between drones are implemented sending signals at short distance over airwaves.
However, standardize and agree on a hardware, a set of frequencies, etc… that could suit the needs all possible collective behaviors has proved to be an illusory goal.
ExaQuark company proposes instead software based solution.
We assume that in a near future every drone should and will be equipped with standard mobile cellular technology giving the drone an internet access over 2G/3G/4G/5G.
ExaQuark provides a scalable cloud platform for developers to imagine, test and implement their solution for UAS collective behaviors. ExaQuark unique distributed technology enables proximity communication by computing and connecting the neighborhoods of each drone, with no limit in the number of drones.
Our proposal is to prototype collective behaviors for fleets to prove the safety and usefulness of our solution. We will focus primarily on one scenario:
Self-structured traffic in corridors
In dense urban environment it is useful to define corridors outside of which the drones are forbidden to fly. In this scenario the drones autonomously navigate the corridor avoiding the other drones. To minimize the risk of collision they segregate by speed: the fastest drones fly in altitude while the slowest ones remain near the ground.
To know the mutual position, they all report their location to exaQuark platform which notify their neighbors. To know their own position with accuracy they use not only their IMU and GPS but also visual landmarks in the corridor.
The prototype development has two stages:
Simulation. The simulation is crucial to determine the parameters of the swarming algorithm. Also, during the simulation the safety of the system will be tested under extreme (simulated) conditions.
In situ test with real drones. A swarm of drones will run following a corridor. To match future conditions, we plan to simultaneously fly thousands of drones in high densities.
- a test corridor with a length in the hundreds of meters range
- a test area on land for the preliminary flights
- alternative collision avoidance techniques
- self structured traffic
- swarms of drones simulation
- development platform for fleets