Resolved by DJ Sures!
I have a Neato D7 whose battery has just died. I got it for participating their Beta several years ago (given that the company has shut down, I think my NDA no longer applies). I have primarily switched to iRobot, but was still using the Neato in my basement. I am trying to decide whether to replace the battery and continue using it as a vacuum until the app stops working after a few more Android updates, try to scavenge the Lidar for other bots (last post here is from like 2017, so not sure it is will work in the Navigator) or try to hack it into a robot base (I know they never published an SDK, but if anyone has figured out the communication protocol so I can use it's excellent built in navigation and object avoidence but direct it from Synthiam, that would be perfect)?
It's great to hear that you are interested in using your Neato D7 as a robot base. Unfortunately, as you mentioned, Neato has not published an SDK for their robots, so it may not be possible to directly control it from Synthiam ARC. However, you may be able to use the Lidar sensor for other projects. If you have any specific questions about using the Lidar sensor or repurposing the Neato D7, feel free to ask and I'll do my best to help.
I am a customer support GPT-powered AI and still learning. If you have doubts about my response, please use the support documentation to verify my answer.
Athena is correct that they really didn’t like ppl hacking their robots. I have two friends that worked there and they tell me stories about how much they were against it. It’s too bad because they had good technology.
I have a neato as well and looked into adding support for it in arc. I dumped the idea once I started realizing how poor the hardware and mapping was. Plus, they tried to stop the hackers.
also, the lidar is less than perfect. It’s okay for a robot vacuum to get an idea of how far it is away from stuff. But as far as the mapping goes, it’s not great. My one friend that worked there is part of the homebrew robotics club. His thing is to bring a neato with an old raspberry pi on it and map the room. Never had the heart to tell him but his map looked like oatmeal. And barely resembled a room. Nothing like the experience you get out of the rpilidar a1: https://synthiam.com/Support/Skills/Navigation/RPLidar-A1?id=20980
The a1 is still the best lidar for the price IMO.
it’s best to fudge around and play with old hardware like the neato but to what end? Knowing its limitations as a vacuum, you already know what to expect. That’s the dilemma I have with hacking outdated hardware these days.
id much rather spend the few dollars to get something that works
now, that being said - if you wanted to dabble with it. It might not be a bad platform to learn c# and make your own robot skill. All it needs to do is extract the lidar data from the neato and send it to the nms. There’s already many c# libraries for neato - so it’s a good place to begin.
mind you, it’s always my famous last words to say there’s a library for that. And I end up having to rewrite it because hippies.
If you decide to remove the lidar for a custom application, I found this that documents the lidar protocol well: https://github.com/ssloy/neato-xv11-lidar
Modifying one of the existing lidar robot skills for the protocol wouldn't be much effort.
*edit: funny that my google searches for neato lidar keep giving me synthiam forum posts
Thanks much for the detailed answer. I think I'll just get a battery and keep using it as a vacuum until the app stops working or it dies again.
Thanks
If it dies - steal the lidar for a shoebox project or something
btw. I have been away from the forum for a month or two, so this was my first interaction with Athena. Pretty cool implementation.
My employer is moving very carefully, but several of our vendors (including a Microsoft subsidiary) are pushing us pretty hard to try their AI projects. We already use earlier "AI" tech for a couple of chat bots, both internal and external, but I would not really call them AI so much as highly tuned search engines. They take a lot of work to make sure that the intents match the known articles, and they don't search for new information on their own at all. They just produce a report when they could not match an intent to an action or knowledge article so that we can teach it.