September 2022
Ishaan Khare received his MS Dual Degree in Electronics and Communication Engineering (ECE). His research work was supervised by Dr. Harikumar K. Here’s a summary of his research work on Collision avoidance for aerial robots operating in the context of Unmanned Aircraft System Traffic Management (UTM):
Unmanned traffic management (UTM) has been gaining popularity these days because we are switching to autonomous navigation for drones in a real-world scenario. Unmanned Aerial Systems (UAS) Traffic Management (UTM) provides a reliable holistic solution for multi-UAV missions including air space design, motion planning, communication, control, geo-fencing, contingency management, etc. Motion planning is one of the key elements involved in UTM. Motion planning includes generating collision-free trajectories from the starting position to the goal position for the UAV in an uncertain and dynamic 3D environment. These trajectories can be planned for many kinds of UAVs, they include holonomic UAVs like a quadrotor or non-holonomic UAVs like fixed-winged aircraft (F W A). Traditionally these collision-free trajectories are planned using methods like velocity obstacles and their variants to avoid obstacles in a 3D space. These obstacles can be static or dynamic whereas the aerial vehicles can also be of various kinds like quadrotor drones or fixed-winged aircraft. So, we start with discussing an improvement on the basic velocity obstacle scheme for a single agent like a quadrotor and an FWA in an uncertain 3D environment. This improvement on the basic velocity obstacle to handle measurement and control noise (uncertainty in the environment) is done by extending it to a probabilistic version called probabilistic versions of Velocity Obstacles (PVO). We explore Probabilistic Inverse Velocity Obstacle (PIVO) as an alternative to PVO for free-flying quadrotor systems. Inverse Velocity Obstacles compute effective controls from a sequence of observations on other agents without the need to access ego state information. As a direct consequence of this, the ego state noise is not entailed in probabilistic formulations bringing in verifiable advantages in the form of reduced path lengths, less conservative maneuvers, and reduced occurrences of stopping/hovering to let others pass. These advantages are vividly tabulated in this thesis, showcasing the efficacy of PIVO as an alternative to probabilistic versions of Velocity Obstacles. In particular, we show the benefits of PIVO over PVO in sample complexity as well as overall trajectory lengths. We also show the efficacy of our probabilistic formulation in handling non-parametric and often multimodal noise distributions. Afterward, we extend our navigation and collision avoidance scheme for multiple drones of the same kind navigating in the same 3D space. Here we propose a predictive optimal control-based autonomous navigation and collision avoidance for multiple fixed-wing aircraft (FWA) or for multiple quadrotor drones that also takes into account the physical constraints of the aircraft. The proposed method is im- plemented in a cooperative framework, analogous to the Reciprocal Velocity Obstacle (RVO) framework for autonomous navigation and collision avoidance for FWA swarm moving in a three-dimensional (3D) space. As a result of predictive optimal control and cooperation, we get less conservative maneuvers, re- duced path lengths, and less deviation from the shortest path when compared to the conventional method without prediction and cooperation. Also, the change in performance with respect to varying prediction horizons is analyzed through numerical simulations. Simulation results are provided, highlighting the advantages of the proposed method when compared to a popular method (Optimal reciprocal collision avoidance) for collision avoidance in 3D for FWA swarm and also against the FGA algorithm (Fast geometric avoidance algorithm). Finally, in a typical UTM scenario, we can encounter multiple vehicles of different kinds (heteroge- neous agents) trying to navigate and avoid collisions in a 3D space. So, we try to formalize a navigation and collision avoidance scheme for this multi-agent heterogeneous system of UAVs.