Meeting Minutes
MRC-II System
Design Review 1 (Part 2)
2:00 – 3:00 PM
Peter Kazanzides completed the Design Review presentation that had been started at the previous meeting on February 12, 2003 (the February 19 meeting was canceled due to snow). This meeting reviewed the Risk Analysis and the High-Level Architecture.
The following comments were made during this presentation:
• For Risk Analysis item #2, another possible method of control would be to perform some type of test (e.g., an accuracy test) to verify that the correct configuration parameters are being used.
• For Risk Analysis item #9 (incorrect joint goals due to software error in trajectory generation), an additional safeguard would be to define a “safety volume” (i.e., a region within which the robot must remain).
• The real-time component of the system should include data logging, similar to the flight data recorder in aircraft. For example, the system could record the real-time state data in a circular buffer so that the most recent information is available for system troubleshooting.
• The currently defined safety response is to stop the motion and/or turn off the motor power. This may not be applicable to all cases. For example, one may wish to automatically retract a needle as part of a safety response.
• There was some discussion regarding the handling of the motion queue during a safety response. One option is to maintain the queue so that it can later be restarted. Another option is to flush the queue and let the application handle it.
