Are there any plans to have this kind of feature? Any Misty dev could choose to be part of the same network where the robots could “learn” or pub/sub to data from each other? I’ve always felt this is an obvious feature a robot should have, yet to my knowledge no one has done it, at least for this type of robot.
Yes, the plan is that eventually the robots could share skills and learned behaviors via the cloud.
It’s also a pretty fun use case for a Misty community developer to implement for all of the other folks – a simple service that might learn about any other robots on the same wifi network and start interacting. Or, even a face/object recognition model that looks for other Misty robots in the environment and, seeing them, start interacting visually, orally, etc… I could see that being a fun project.
@station do you see practical applications for that capability within an office or home environment? Or mostly entertaining applications?
I also think it would be tons of fun if there was some communication that could only take place face-to-face between mistys. I bring mine over to your place and they bleep-blorp at each other for a bit, and I after I leave your robot does some new trick it learned from mine.
@tenwall I was looking at stuff like Satori and thinking how connected and real-time things can become so almost having each Misty robot be contributing or acting as singular larger brain or taping into a larger brain to learn information. And then perhaps useful information could be propagated to others, not just for entertainment and home but anywhere. a camera, internet connection and mobility can do a LOT these days. that being said, I feel the info needs to be accessible outside a walled “cloud” to be truly useful.
@Dan I think those type of unique one on one interactions are fun as well and require the extra social interaction for the payoff which I think is a good nudge to connect devs.
@station now I understand. We have talked a lot about how “the community” of robots could contribute to each others’ knowledge. One of the examples we use is:
Robot 1: “my owner just taught me what a coffee cup looks like”
Robot 2: “so did mine”
Robot 3: “funny - my owner told me [this screwdriver] was a coffee cup; that must not be true”
…and so on. In that manner, the amount of learning on an individual robot level could be exponential.
That, of course, is not a v1 type of capability. Baby steps first; then robots teaching each other.