Blog 2018-08-08

Github Site and Updates

2018-08-08

The big news is that the Github repository for the Caveatron is now public! There you will find the software that runs the system along with documentation including a bill of materials, drawings, wiring diagrams, and instructions on how to go about building one. If you want to contribute, please contact me and I can add you as a member. I still need to prepare a user manual and I need to clean up the desktop computer software used for post processing a bit before they are posted for download. Those will be available from this website, which will focus on use of the system, whereas the Github repository will be focused on building it. Furthermore, there is also a Google group about all aspects of the Caveatron at caveatron-discuss.

A new update to the code was just completed adding a new feature I have wanted to include for a while which is the ability to view LIDAR scan data onboard after a scan is completed. Due to limitations of the Caveatron processors it’s not possible to create an interactive or 3D view of the data, so the most useful option I could think of was to create a horizontal and vertical slice through the data points to produce a rudimentary plan and profile view of the scanned area as green dots. The plot also includes a representation of the path the user took while creating the scan with each yellow dot being where a valid LRF distance was obtained. Simple zoom controls and a scale bar are also included.

Screenshots from the new LIDAR data review screens showing plan and profile views of a cave passage.

Other than possible bug-fixes I only have one more update planned for a while and that is integrating a new LIDAR scanner that looks very promising, the RPLIDAR by SLAMTEC. Several RPLIDAR models are available with ranges from 12 m to 25 m and with different scanning rates, but they all use the same protocol. The A1M8 is especially interesting since it can serve as a complete replacement for the XV11 LIDAR, but with 3 times the range and potentially nearly 5 times the scan rate. It is essentially the same form factor, works just as well as short distances, and best of all costs only a few dollars more than you can find an XV11 LIDAR on Ebay! Early testing shows that it should meet its rated 12 m range just fine even at its maximum 8 kHz sampling rate, but it looks like integrating it with the Arduino to work at a data rate faster than 2 kHz will be a complicated undertaking. Even at 2 kHz though, its still faster than either the XV11 or SWEEP LIDARs.

After that, I plan to shift from software to developing a 3D printable enclosure design to simplify fabrication and also implement an idea I have for a better way to attach the LIDAR modules.