Mitchell Update 4/7 – 4/13

Mitchell Update 4/7 – 4/13

Sensor Issues

This week I ran into a bit of a hurdle with the sensors. I would seemingly randomly get consistently false data (MAX_RANGE_VALUE) from the sensors that could be removed from restarting. It seemed to come from just one sensor, but to make it more difficult the port a sensor is plugged into does not correlate to the scanner port in the code. Ultimately I found that two sensors were stuck in configuarion mode and doing a factory reset I was able to get one working again. For now I am using the backup sensor to replace the fault one and we have ordered more.

Code to configure modes of lidar sensors on setup LiDAR

Wrote code to set the sensor mode (long, mid, short, dynamic). By default, the sensor is in dynamic, which means it changes based on received values, but there is a latency delay to switch. Since I know we only care about data from 3-8+ meters, we can just use long range mode. Most of the other default setting match our needs so I did not configure them.

Safe bootup of lidars

I worked on writing code to safely boot the pi and the sensors. This includes attempting to retry connecting/configuring sensors on bootup, and timing out on failure and notifying the user as such. This makes it easier for the demo, so that we can quickly find a faulty sensor or connection in our prototype.

Integrate advanced lidar incoming object detection algorithm

I worked on integrating a more advanced algorithm for incoming object detection. I wrote the outline that would include, filtering outliers, filtering an object that the user is passing by. Figuring the speed at which it is approaching (placing into categories of immediate, moderate, slow) (could be used for better integration with our pipeline).

Next week

Finished advanced incoming object detection

Integrate with the rest of the group

Debug and improve code

 

Leave a Reply

Your email address will not be published. Required fields are marked *