Figure 1: Schematic of the simulation software that individual investigators will be collaborating with each other to build.
Collaboration on SmartPATH simulation:
Faculty involved:
A. Agogino
K. Hedrick
J. Malik
M. Tomizuka
P. Varaiya
Figure 1: Schematic of the simulation software that individual investigators
will be collaborating with each other to build.
A common goal shared by Profs. Varaiya, Malik, Tomizuka, Hedrick, and Agogino is to link the models for the various sensors (including vision), the vehicle dynamics, with the control algorithms, and interface it with the SmartPATH simulator. In order to achieve this goal the various investigators will be communicating and interacting with each other. The common hope is that the interaction and communication between the independent research groups will enable them to build a realistic simulation package that will allow the different investigators to test their methodologies and control algorithms on comprehensive sensor and vehicle models.
Figure 1 shows the modular approach taken for this task. As shown, the different modules consist of the sensor models (Prof. Agogino), the vision sensor (Prof. Malik), the sensor validation, fusion and diagnosis (Prof. Agogino), the car model (Profs. Tomizuka and Hedrick), the vehicle level controllers (Prof. Hedrick and Tomizuka), and the actual SmartPATH simulation module (Prof. Varaiya). In implementation a cycle will be created consisting of 1.) generate sensor readings through sensor models, 2.) fuse readings, 3.) control, 4.) simulate vehicle dynamics, 5.) show results in SmartPATH, 6.) input this to sensors. The sensor models use characteristics from the various sensors obtained experimentally and analytically and model the uncertainty caused by noise and failures. The sensor fusion and diagnosis module validates the sensor readings and performs diagnosis. The signals are then fed into the longitudinal and lateral controllers which act at the vehicle level of the vehicle. This will result in an action obtained by using the car models which will be computed in the SmartPATH simulation module to generate an image of the scene. The scene image will be computed twice from viewpoints slightly apart and elevated to simulate the stereo vision cameras. These images will then be fed into the vision sensor module to allow for an integration of this sensor as well. This use of SmartPATH does not follow the approach of simulating a whole sequence of actions before displaying them but uses a frame by frame approach. The information from the vision sensor is then fed into the fusion module where it meets the readings from the other sensors to close the simulation loop. The architecture is structured in a modular manner which allows to replace or add other modules.
References
Agogino
Alag, S., K. Goebel, A. Agogino, " A Framework for Intelligent Sensor Validation, Sensor Fusion, and Supervisory Control of Automated Vehicles in IVHS", accepted for presentation and publication in the conference proceedings of the ITS America 1995 conference, Washington, D.C. March 15-17, 1995a.
Alag, S., K. Goebel, A. Agogino, " A Methodology for Intelligent Sensor Validation and Fusion used in Tracking and Avoidance of Objects for Automated Vehicles", accepted for presentation and publication in the conference proceedings of the ACC 1995 conference, Seattle, WA, June 1995b.
Hedrick
Hedrick, J.K. and Garg, V., "Failure Detection and Fault Tolerant Controller Design for Vehicle Control," PATH Research Report draft no. 93-09, 1993.
Hedrick, J. K., McMahon, D., Swaroop, D., Garg, V., Gerdes, J., Maciuca, D., Blackman, T., Yip, P, "Longitudinal Control Development for IVHS Fully Automated and Semi-Automated Systems-Phase 1", California PATH Research Report, UCB-ITS-PRR-95-4, January 1995.
Malik
Koller, D., Weber, J., Huang, T., Malik, J., Ogasawara, G., Rao, B., and Russell, S., "Towards Robust Automatic Traffic Scene Analysis in Real-Time", Proceedings of the International Conference on Pattern Recognition, Jerusalem, October 1994.
Koller, D., Luong, T. and Malik, J., "Binocular Stereopsis and Lane Marker Flow for Vehicle Navigation: Lateral and Longitudinal Control, Report No. UCB/CSD 94-804, March 24, 1994.
Tomizuka
Chee, W. S. and Tomizuka, M., "Vehicle Lane Change Maneuver in Automated Highway Systems, California PATH Research Report, UCB-ITS-PRR-94-22.
Patwardhan, S., Tomizuka, M., Zhang, W-B, and Devlin, P., "Theory and Experiments of Tire Blow-Out Effects and Hazard Reduction Control for Automated Vehicle Lateral Control Systems,", Proceedings of the International Symposium on Advanced Vehicle Control, Tsukuba, Japan, Oct. 1994.