This week I finally obtained the parts on Tuesday. On Wednesday during the remote class work session, I borrowed my roommates monitor and keyboard to set up ssh on my raspberry pi. Since I do not currently own a monitor and keyboard, I will be using ssh until the monitor and keyboard we ordered arrives.

I later spent a few hours trying to install cartographer onto the raspberry Pi. I was unable to do so successfully. In our remote work session on Thursday, I worked with the rest of the team for a couple of hours to try to install cartographer. Since we were finding it extremely difficult to do so, we decided to use BreezySlam – a python SLAM library.

I tested out this library on my macbook with a lidar. The results were not so good – the resulting map was not an accurate depiction the the lidar scan, and without odometry data from the robot, the SLAM program kept crashing in it’s attempt to calculate the position of the robot.

In our work session on session on Saturday, I assembled the robot using tape:

I also used another python library called BreezyCreate2 to control the roomba. It is now possible to control the robot remotely over ssh as seen in this video:

https://drive.google.com/file/d/1bW5QiCLh0HIJ45aBB3qgRIBUMouuUv-K/view?usp=sharing

While I’m pretty happy with the progress so far, one possible technical difficulty is that it is a little difficult to view slam/lidar scan results over ssh. I’m using x forwarding over ssh, and the lidar display generated by matplotlib is a little laggy.

Next week I plan to address this, as well as integrating roomba odometry into the sample RPlidar slam program provided by BreezySlam.