Working on a new robot controller as time allows. It’s based on a cardputer adv which is part of the M5 ecosystem. Just ordered a tiny analogue joystick to use with it and considering exploring LoRa communications. Found that current Bluetooth control is limited by distance and obstacles, LoRa should work better in both regards. […]
Back to looking at the BrainChip neuromorphic computer. Ran into software issues previously which need revisiting but here testing a potential (prototype) portable power system. There is a solar panel option underneath the keyboard/display. However, as the sun is somewhat reluctant to shine at times this is using a USB power bank. The LED display
Neuromorphic Field Kit Read More »
This device crawls at the moment. The rear wheels are, for now, freewheeling so they’re dragged. Imagining the front ‘whiskas’ additionally detecting something or other as they sweep from side to side. It is a hacked DFrobot Insectbot. Yellow washers are ballast (big buggy ‘eyes’ are sensors). No software updates just yet so it still
The Mother of Invention Read More »
Still planning (for want of a better term) the neuromorphic Raspberry Pi 5 so that it can be powered by solar energy. Decided to add the battery gauge as well as external LEDs for indicating charge/done, etc by both solar and other energy sources. As you can see this is an extremely early mock up
The above video is the robots executing the demonstration code which was executed remotely. It will be a relatively steep learning curve now to try to replicate the detection swarming behaviour described elsewhere on this site. This can then be executed remotely too then no doubt refined to try to work towards proving or refuting
Very early experiment in concurrent control of multiple elements: a tracked robot, a robot dog and a drone. It’s taken some effort to get this far and there’s still further to go! Programmatic control requires improvements. Here the cognitive burden is quite high although with practice that should lessen. Additionally, there is no clear shared
Better view of the scene below using a Raspberry Pi Camera 3 Module with a wide Field of Vision (FoV). Good auto focus too. Again, this is a snapshot of the video streamed from the Pi on the ceiling of a small room to the ‘solar computer’. Dimensions now are reasonable at approximately 260cm x
Experimental Parameters The Sequel Read More »
The above is a screenshot of a ‘birds eye’ video of an area of interest below. The area is approximately 80cm by 104cm and the height of the camera is roughly 240cm above the centre of the area. The general feeling is that if this is used ‘as is’ to begin to demonstrate ‘swarming’ of
Experimental Parameters Read More »
Thinking about having a Big Brain outside of the robots. The Little Brains onboard can handle local behaviours, inputs and outputs from sensors to actuators respectively, whilst Big Brain does more serious computing such as image processing, direction finding, swarm coordination etc. It’s not too challenging technically as the ‘solar computer’ already talks to the
