Matthew Turpin

Trajectory design and control for aggressive formation flight with quadrotors

In this work we consider the problem of controlling a team of micro-aerial vehicles moving quickly through a three-dimensional environment while maintaining a tight formation. The formation is specified by shape vectors which prescribe the relative separations and bearings between the robots. To maintain the desired shape, each robot plans its trajectory independently based on its local information of other robot plans and estimates of states of other robots in the team. We explore the interaction between nonlinear decentralized controllers, the fourth-order dynamics of the individual robots, time delays in the network, and the effects of communication failures on system performance. Simulations as well as an experimental evaluation of our approach on a team of quadrotors suggests that suitable performance is maintained as the formation motions become increasingly aggressive and as communication degrades.

Decentralized formation control with variable shapes for aerial robots

We address formation control for a team of quadrotors in which the robots follow a specified group trajectory while safely changing the shape of the formation according to specifications. The formation is prescribed by shape vectors which dictate the relative separations and bearings between the robots, while the group trajectory is specified as the desired trajectory of a leader or a virtual robot in the group. Each robot plans its trajectory independently based on its local information of neighboring robots which includes both the neighbors planned trajectory and an estimate of its state. We show that the decentralized trajectory planners (a) result in consensus on the planned trajectory for predefined shapes and (b) achieve safe reconfiguration when changing shapes.

Capt: Concurrent assignment and planning of trajectories for multiple robots

In this paper, we consider the problem of concurrent assignment and planning of trajectories (which we denote Capt) for a team of robots. This problem involves simultaneously addressing two challenges: (1) the combinatorially complex problem of finding a suitable assignment of robots to goal locations, and (2) the generation of collision-free, time parameterized trajectories for every robot. We consider the Capt problem for unlabeled (interchangeable) robots and propose algorithmic solutions to two variations of the Capt problem. The first algorithm, c-Capt, is a provably correct, complete, centralized algorithm which guarantees collision-free optimal solutions to the Capt problem in an obstacle-free environment. To achieve these strong claims, c-Capt exploits the synergy obtained by combining the two subproblems of assignment and trajectory generation to provide computationally tractable solutions for large numbers of robots. We then propose a decentralized solution to the Capt problem through d-Capt, a decentralized algorithm that provides suboptimal results compared to c-Capt. We illustrate the algorithms and resulting performance through simulation and experimentation.

Concurrent assignment and planning of trajectories for large teams of interchangeable robots

This paper considers the problem of finding optimal time parameterized trajectories for N unlabeled robots navigating through a cluttered environment to N unlabeled goal locations where success is defined as every goal being reached by any robot. We propose a complete computationally-tractable algorithm for simultaneously finding trajectories and assignment of goal locations. This method is then demonstrated to have an upper complexity bound of that scales polynomially in the number of robots, O(N3). The trajectories generated are guaranteed to be minimum length and collision free, while the assignment policy minimizes the maximum distance travelled. The key idea in the paper comes from the coupling between the optimal assignment, the properties of the resulting paths, and the set of valid priority assignments to the robots. These benefits result from structure in the solution to the optimal assignment to create a partial ordering of the robots, which in turn allows safe trajectories to be easily generated. Finally, we demonstrate the performance of the algorithm through simulations with tens and hundreds of robots operating in cluttered and confined environments.

Goal Assignment and Trajectory Planning for Large Teams of Aerial Robots.

This paper presents a computationally tractable, resolution-complete algorithm for generating dynamically feasible trajectories for N interchangeable (identical) aerial robots navigating through cluttered known environments to M goal states. This is achieved by assigning the robots to goal states while concurrently planning the trajectories for all robots. The algorithm minimizes the maximum cost over all robot trajectories. The computational complexity of this algorithm is shown to be cubic in the number of robots, substantially better than the expected exponential complexity associated with planning in the joint state space and the assignment of goals to robots. This algorithm can be used to plan motions and goals for tens of aerial robots, each in a 12-dimensional state space. Finally, experimental trials are conducted with a team of six quadrotor robots navigating in a constrained three-dimensional environment.

Goal assignment and trajectory planning for large teams of interchangeable robots

This paper presents Goal Assignment and Planning: a computationally tractable, complete algorithm for generating dynamically feasible trajectories for

Decentralized controllers for perimeter surveillance with teams of aerial robots

N

Self-assembly of a swarm of autonomous boats into floating structures

N interchangeable (identical) robots navigating through known cluttered environments to