The major risks to our project at this point can mostly be traced back to time. We don’t foresee any critical roadblocks at this point, but getting everything up and running and integrated will be challenging. In our meeting, we planned out a few strategies to modularize the project as much as possible, so that work can be done in parallel. Currently, we separated the project pretty cleanly into three parts, with each of us working offline, then meeting to continue work and integrate.

No major changes have been made since we switched to BreezySLAM over Cartographer. The basic internal program structure has been implemented, and individual subsystems are already working for basic test cases. This week we focused on tweaking the SLAM algorithm and connecting it to odometry data to yield the best map possible.

Our main goal has been to generate concrete deliverables before our work sessions, and before the demo. These include:

• Move robot autonomously
• Read LIDAR data from Pi
• Process SLAM output for use by our algorithms
• Basic navigation:
  • Breadth-first path planning to choose a destination based on some criteria
  • Shortest-path route planning to identify how to move to the chosen destination