I have a situation at the moment where some kind of remote controlled eye would be super handy. So I got to thinking, how could I build a teeny weeny little camera + locomotion + telemetry that was small as possible while being able to manage weird terrain. An ESP8266 with some kind of small camera? I found this, but even it is too big I think. Some sort of micro battery? Runtime wouldn’t have to be long, maybe one of those teeny 8g lipos. I’m totally stuck on ideas for locomotion, I just cant think of anything simple that would be able to deal.
But its kind of a fun mental exercise, no? How would you build the smallest, most mobile camera possible? You could wrap a flexible pcb around a battery core, and put a camera on the end! that would be pretty good. You could jam the thing onto the smallest rc car you could find, but even they are pretty big. What if you could constrain the method of movement, ie by attaching the eye to a cable or pipe to run along? I’d probably have it drag a string for retrieval in any case. Maybe you could make it hover somehow and bounce its way along? Or squidge its way along by utilising one of those tubes made of gel…
Those baby quadrocopters like James brought to the space seem to fit the bill for this… the on-board camera isn’t spectacular but then again its probably fine so security type needs. On the other hand that may be harder to automate if that’s your desire.
Alternately if your wanting near silent operation or longer run times you could hang something from a wire. A couple of concave wheels sitting on top of the cable would be all it needs for locomotion, so the wire could be supported from the bottom allowing for cheap\easy installation. The other advantage of this kind of prelaid track based system is with some careful planning you wouldn’t even need any pan or tilt control on the robot, it’ll always point away from the wall so preset the tilt, and just put a half circle bump in the track at occasional intervals so as you roll around it you get a pan. This limits you movement to a single DC motor, and if your not needing it to ever go very fast a simple pair of small relays\mosphets is adequate to control it. It also makes it easy to have a basic two contact charging dock at one end of the track for it to home to. The drawback is however, also just as simple, if there is no track in an area, you can’t get video.
Haha, some interesting ideas lads but we’re talking on different scales I believe. You’re thinking small as in the palm of your hand, I’m thinking small as in perhaps a tube 15mm diameter or so (length relatively unimportant :P) Basically, if the camera lense is the largest component ( bulk or diameter wise ), we’ve done exceedingly well
Also, should probably reiterate, not a practical, I’m actually going to build this thing. Not unless we come up with some amazing idea. Just curious as to how others would solve the problem. I’m currently trying to figure out if you could do reasonable ( controllable ) locomotion via vibration, and take camera frames in between pulses.
Also! Consider the point that the parameter is an uncontrolled track. Think exploring caves or wall cavities. Hence the ‘how do you get around stuff when you’re that little’ conundrum. Or the alternative locomotion of latching to an existing cable or line perhaps. Or locomoting through a pipe.
Heres an idea, maybe you could cover it in polarised magnetic balls that are sticky, and you could rotate the balls with magnetic pulses from underneath to create a sort of undulating concept. Similar to the mechanism behind e-ink.
I’m not to sure where you want to stick that thing in.
But if that should fit the bill https://www.youtube.com/watch?v=ISznqY3kESI
and you are keen to fund the project I do the CAD/printing.
Wanted to build one of these since ages, looks like here is an actual reason to do it.
you’d probably want to gear that down but oh my christ trying to source gears that tiny may be literally impossible. Watch gears perhaps? do watches use many bevels?
Plus the motors above, plus a battery, a smattering of standard rc circuitry and that would build a pretty interesting, seriously little micro walkabout wireless eye. video server and adaptor optional, just depends what you want to see the output on.
I have been thinking about hosting a hack a toy evening/competition. Think “Furby your plush”.
I have a cuddly turtle looking plush toy (once owned by a dog for a day before it was harshly rejected) I wanted to convert into a camera’d moving wireless network controlled turtle. Only because its locomotion on land would be ridiculous and well robots without a purpose are hilarious.
