PS - Apologies for not being very active lately. It’s been a busy time - I just accepted a new job, and will be moving in a couple of months. So may continue to be more lurker than I’d prefer.
By then I hope to have the GoPiGo3 encoders /odom EKF fused with the /imu/data using robot_localization, AND have the Oak-D-Lite doing 3D depth obstacle detection.
The only thing Turtlebot 4 iRobot version will have on the GoPiGo3 Dave at that point will be that silly navigate to the dock and recharge thing.
After “I” learn ROS2 localization, ROS2 mapping, ROS2 navigation, ROS2 3D depth obstacle detection, and figure out how to run all that on the Pi3B+, Dave should be well on his way to finding Carl’s dock, recognizing that Carl is not using it at that moment, and claim squatter’s right to a recharge.
Realized I better buy a set of original iRobot charging contacts for Dave while they are still available.
I’ll still need to get the 2A 11.1v (3S) Li-Ion charger from Tenergy $19 and buy Carl a TalentCell pack, and change my rule-based charge/trickle detection algorithm for the Li-Ion technology.
Oh boy, can you imagine the fight over who needs a charge more - Dave or Carl?
Wasn’t that long ago you considered a lithium battery for Carl high treason!
You better teach Carl not to sit on Dave by mistake too.
How will they tell their docks apart?
Instead of going to all that trouble, (inventing a whole new OS paradigm and all ), how about a beacon instead - and a foam-core guide that will help him align with the dock when he gets close enough?
Dave has “survived” 196 playtime/recharge sessions including 8 safety shutdowns without incident. All the charges have been at 1A max, “attended”, and carefully watched for the green “full charge” indication.
I trust the Tenergy smart chargers, but charging the TalentCell battery at 2A will need to achieve an equal track record before bringing another bot onto the Li-Ion chemistry.
(I submitted an iRobot Create 3 beta tester application tonight. The Create 2 is only $200 but the Create 3 is ROS2 based. Slap a PI-4 and the Oak-D-Lite on top of it - may be a winner. (albeit it won’t win the tiny robot award.)
BTW: with respect to my “I’ve never owned a robot that didn’t have a wall following example”:
I would follow the profile used by the TalentCell charger - and I’m not sure about the 2 amps. I wouldn’t go higher than one amp, possibly less based on the TalentCell charger’s max.
And where are you going to put all these beasties?
So. . . .
What prevents you from adding one to either of your existing robots? I even provided schematics!
When I was a teenager, (that had never lived in a single-family home), I designed a pentagonal home with a central great room.
After 35 years of homes, I finally found a great room home that did not have the wasteful living room/ family room redundancy. It does have wasteful redundant dining areas, one containing my piece of art dining table to welcome all the guest robots that come to juice-up around the table.
I’ve actually put a lot of thought into this subject and concluded the GoPiGo3 needs a 3D printing facelift to remain competitive in the educational robot market. It is consistently a fantastic bot, with superb learning material and support, but the image is not “refined”, “modern”,or “latest tech” to the untrained observer. No other successful robot is running around the room “naked”
(Well perhaps it could be said the TurtleBot 3 is also naked and successful)