[HN Gopher] Capturing robot data so we can debug it later
___________________________________________________________________
Capturing robot data so we can debug it later
Author : davidtarazi
Score : 15 points
Date : 2022-10-18 16:04 UTC (6 hours ago)
(HTM) web link (www.polymathrobotics.com)
(TXT) w3m dump (www.polymathrobotics.com)
| luc_ wrote:
| Nice... I've built a very similar toolset in the past. Cool to
| see it open-sourced. Nice work.
| davidtarazi wrote:
| Thanks! It's definitely a common problem for robots and I hope
| others can benefit from this.
| kortex wrote:
| Nice that the ROS community is starting to get hip to web/cloud
| community innovations. Back when I worked on ROS heavily (about 2
| years ago), it felt like there were tons of cases of Not Invented
| Here syndrome. Rosparam server, service discovery, some of the
| rpc concepts. Just felt like a large chunk of the stack could be
| replaced with pre-existing technologies that could do the job
| better and with more features.
|
| TFA is traditionally the sort of thing someone would write a full
| solution for, entirely in ROS, and it would be rickety and
| opinionated, but refreshingly, they just use Prometheus, with
| enough Ros binding to capture the auxiliary data.
| davidtarazi wrote:
| Yup, I completely agree. I'm fairly new to the ROS community,
| but even so a lot of the solutions that I've seen for various
| robotics problems could be solved with cloud/web innovations.
| Never reinvent the wheel!
___________________________________________________________________
(page generated 2022-10-18 23:03 UTC)