MIT Uses Adams to Simulate Mobile Robotic Systems
The Robotic Mobility Group at MIT has extensive experience designing, modeling, and controlling mobile robotic systems on uneven and rough terrain, with applications involving planetary exploration rovers, autonomous military vehicles, and automotive passenger vehicles.
High-fidelity simulation of robotic systems with Adams is a valuable research tool for our lab. By facilitating virtual prototyping of robot mechanisms and controllers, it allows us to maximize the value of our physical experiments. For example, we used ADAMS simulations as virtual prototyping of a vehicle stability measure that we were later able to demonstrate experimentally (see Stability Measurement of High Speed Vehicles).
Adams has a number of useful features that improve our research productivity. Models can be built directly in Adams or imported from an external CAD package. For automotive applications, the Adams/Car package includes numerous vehicle model templates and tire models that can be tailored to a particular vehicle. Once the 3D model is in place, it automatically considers dynamic effects such as gyroscopic torques, wheel slip, uneven terrain, and vehicle rollover. Coding these effects from scratch can take a lot of time.
Note how the tire forces shift during this abrupt turning maneuver.
With the Adams/Controls plugin, we are able to connect our robot models to Matlab and Simulink to simulate a closed-loop control system. This allows us to use existing Matlab and Simulink code in conjunction with an Adams model. We have used this feature to develop control algorithms for autonomous and semi-autonomous avoidance of both stationary and moving hazards (see animations). Hazard avoidance involves aggressive maneuvers in which many dynamic effects come into play. Using ADAMS for our model helps to pinpoint the problems with simpler models before experiments and allows us to be more confident in our simulation results.
Finally, Adams has a number of useful debugging features; including a built-in plotting and animation tool that really help us work out kinks in our simulation models. Being able to watch an animation of the model and put sensors anywhere in the system makes it much faster to diagnose problems as they arise.
Overlaid results of two vehicle simulations - for flat road vs rough road.
We have found Adams to be a useful tool for our research.
Peters., S., and Iagnemma, K., "Mobile robot path tracking of aggressive maneuvers on sloped terrain," Proceedings of the IEEE International Conference on Robots and Systems, IROS '08, 2008.
Peters, S., and Iagnemma, K., "Stability Measurement of High Speed Vehicles," Vehicle System Dynamics, v. 47, n. 6, pp. 701-720, June 2009.
Anderson, Sterling J, "A Unified Framework for Trajectory Planning, Threat Assessment, and Semi-Autonomous Control of Passenger Vehicles," Master's thesis, Massachusetts Institute of Technology. Dept. of Mechanical Engineering, 77 Massachusetts Avenue, Cambridge, MA 02139, June 2009.