SunSeeker MkI

V0.1 (2014-07-06)

Well, I suppose I've made a bit of progress. More issues than anything. First, something in my arduino sketch seems to be conflicting with the servo, since it just twitches. I eventually gave up trying to debug this, and programmed it to navigate without the servo. It's currently detached. You can see how well this works in the video (not very). Some things bear explaination: You'll notice the robot stops every second. This is because running the motors seems to affect the rangefinder, so we have to stop to get an accurate reading. You'll also notice we get stuck against a cardboard box at the end, this is because its in our blindspot. It wouldn't be an issue if I could get the servo working...

The second major thing I did was add a display (this one, specifically). I want to be able to show different faces depending on the program status, to give the 'bot a little more personality (see the photo). I was able to get it working right away, which was nice, and somewhat unexpected. Unfortunately, every silver lining has its cloud. It seems the UNO doesn't have enough SRAM to run my control program and manage the display simultaneously.

At this point, I'm not really sure where to go with this project. I could solve my issues with hardware; a controller for the servo and a more powerful MCU for the display, but that would mean more power consumption and more money. Not to mention, if I really want to deploy this outdoors, I'll need a more robust chassis.

V0.1 (2014-06-12)

The SunSeeker, as the name implies, is designed to seek out sunlight to charge the batteries. The idea is, I once I make it self-sustaining, I can add more useful functionality, and put it in a chassis more suited to the outdoors.

• Control: Arduino Uno and a TB6612FNG motor driver
• Sensors: HC-SR04 ultrasonic rangefinder, and an ACS714 current sensor
• Power: 2x 9V solar panels in parallel and 6x rechargeable NiMH AA batteries
• Mechanics: Mini servo for moving rangefinder, and an RC Ford F-150 as the chassis

I'm interested in learning about different methods of control. Right now, I'm studying behaviour trees. Currently, the programming just tests the motors and checks the power system.