Charting a safe course through a highly uncertain environment

An autonomous spacecraft exploring the distant reaches of the Universe descends through the atmosphere of a remote exoplanet. The vehicle, and the researchers who programmed it, don’t know much about this environment.

With so much uncertainty, how can the spacecraft plot a trajectory that ensures it won’t be crushed by any moving obstacle or blown off course by sudden, gale-force winds?

MIT researchers have developed a technique that will allow this spacecraft to land safely. Their approach can enable an autonomous vehicle to plot a demonstrably safe trajectory in highly uncertain situations where there are multiple uncertainties regarding the environmental conditions and objects with which the vehicle could collide.

The technique can help a vehicle find a safe course around obstacles that move in random ways and change shape over time. It calculates a safe trajectory to a specific area even when the vehicle’s starting point is not exactly known and when it is unclear exactly how the vehicle will move due to environmental disturbances such as wind, ocean currents or rough terrain.

This is the first technique to address the problem of trajectory planning with many concurrent uncertainties and complex security constraints, said co-lead author Weiqiao Han, a graduate student in the Department of Electrical Engineering and Computer Science and the Computer Science and Artificial Intelligence Laboratory (CSAIL).

“Future robotic space missions need risk-aware autonomy to explore remote and extreme worlds for which only highly uncertain prior knowledge exists. To achieve this, trajectory planning algorithms must reason about uncertainties and deal with complex uncertain models and safety constraints,” adds co-lead author Ashkan Jasour. , a former CSAIL researcher who now works on robotic systems at NASA Jet Propulsion Laboratory.

In addition to Han and Jasour on the paper is senior author Brian Williams, aerospace professor and member of CSAIL. The research will be presented at the IEEE International Conference on Robotics and Automation and has been nominated for the Outstanding Paper Award.

Avoid assumptions

Because this trajectory planning problem is so complex, other methods of finding a safe way forward make assumptions about the vehicle, obstacles and the environment. These methods are too simplistic to apply in most real-world situations, so they can’t guarantee that their trajectories are safe in the presence of complex uncertain security constraints, Jasour says.

“This uncertainty could be due to the randomness of nature or even the inaccuracy in the autonomous vehicle’s perception system,” Han added.

Instead of guessing the exact environmental conditions and locations of obstacles, the algorithm developed reasons about the likelihood of observing different environmental conditions and obstacles in different locations. It would make these calculations using a map or images of the environment from the robot’s observation system.

Using this approach, their algorithms formulate trajectory planning as a probabilistic optimization problem. This is a mathematical programming framework that allows the robot to achieve planning goals, such as maximizing speed or minimizing fuel consumption, while taking into account safety constraints, such as avoiding obstacles. The probabilistic algorithms they developed reason about risk, which is the likelihood of not meeting those security constraints and planning goals, Jasour says.

But because the problem involves several uncertain models and constraints, from the location and shape of each obstacle to the starting location and behavior of the robot, this probabilistic optimization is too complex to solve with standard methods. The researchers used higher-order statistics of probability distributions of the uncertainties to convert that probabilistic optimization into a more straightforward, simpler deterministic optimization problem that can be efficiently solved with existing out-of-the-box solvers.

“Our challenge was how to reduce the size of the optimization and consider more practical constraints to make it work. It took a lot of effort to go from good theory to good application,” says Jasour.

The optimization solver generates a risk-based trajectory, which means that if the robot follows the path, the probability that it will collide with an obstacle does not exceed a certain threshold, such as 1 percent. From this, they obtain a series of control inputs that can safely steer the vehicle to its target area.

Mapping courses

They evaluated the technique using several simulated navigation scenarios. In one model, they modeled an underwater vehicle that charts a course from an uncertain position, around some odd-shaped obstacles, to a target area. It was able to reach the target safely at least 99 percent of the time. They also used it to map out a safe trajectory for an air vehicle that avoided various flying 3D objects that have uncertain sizes and positions and could move over time, while in the presence of strong winds their movement could be reduced. was affected. With the help of their system, the aircraft most likely reached its target area.

Depending on the complexity of the environment, the algorithms took a few seconds to a few minutes to develop a secure trajectory.

The researchers are now working on more efficient processes that would significantly reduce runtime, bringing them closer to real-time scheduling scenarios, Jasour says.

Han is also developing feedback controllers to apply to the system, which would help the vehicle stay closer to its planned trajectory, even if it sometimes deviates from the optimal course. He is also working on a hardware implementation with which the researchers can demonstrate their technique in a real robot.

This research was supported in part by Boeing.

#Charting #safe #highly #uncertain #environment

Leave a Comment

Your email address will not be published. Required fields are marked *