BEAM navigation

Phototropic_2m_walker.gif

This is the circuit for a phototropic walking robot insect which follows light.
It's based on M.Tilden's work, BEAM robotics: it's about analog control circuits (microcores) that simulate neural and nervous networks present in the in the real insects' spinal cord; these react to external stimuli in a robust and very reliable way. Many other circuits have been drawn to do things like follow light, sense obstacles and reverse motors.
Of course this works just for walkers for the moment.

here comes my project: a BEAM autopilot!
It will be based on conversion of digital data coming from sensors, GPS to analog current levels, that will be inputted to control circuitries like the one above. The outputs will then be reconverted to digital for actuating servos and so on. I haven't still drawn nor calculated anything.

This is just a newborn concept, now I'd like to hear opinions of feasibility from experts in electronics I'm such a stinky newbie!

Thanks to everyone who will give me some thoughts!!
E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • was just thinking it could be a funny thesis project for me. When I have some spare time i'll draw down some schemes goooodbye to evribadi
  • Interesting idea... but it sort of seems like Rube Goldberg way of flying. Going analog leaves you open to all sorts of new problems. If your main concern is robustness, it wouldnt be too hard to add a watchdog timer to your digital AP and have a backup AP that can take over if it locks up. Or there's the good old switching to remote control plan. I really can't see analog being more practical.

    On the other hand if you're in it for the journey, sounds like quite the unique and entertaining project. Keep us posted!
  • What i dpn't like about digital electronics is that systems get very complicated and can crash or glitch easily (think of windows vista) : during a flight, this would mean destruction. I tried to build a BEAM bot and it works really great and fast, and it can't lock up. Can be interesting, if done, but it's such a hard task to accomplish!
  • if you want a bio-inspired auto-pilot you could do something with neural networks in software.
    I am not going to get into why your idea is not feasible because I don't know where to start.
This reply was deleted.