Making progress

After a couple of frustrating weeks we are starting to get to grips with this robot. The trick, for us at least, is to keep it simple. Last week we had some difficulty with the line following programme. The robot could not operate in the ambient light in the room, it was chucking it down outside and the light in the room was very dim. The result of this was that the robot could not 'see' the green line it was looking for and went off on a random course of its own as it circled round to try and find a green line so it could turn onto the transition table. The result was a devastated table and some puzzled programmers. The dim light also meant that the robot picked up a grey line between base and the gardening patch, thought it was green and turned straight into the woodworking table.

All in all it was a bit of a disaster but it resulted in an important decision to only use the light sensors when absolutely necessary. This might seem obvious but the decision has important consequences. Using the cues on the field mat means you reduce the errors in your programming, so we will loose a good way of keeping the robot on track. It is possible to calibrate the light sensors so the robot can adjust to the ambient light conditions however we are having a little trouble learning how to do this consistently. Sam will still be doing some complex line following for the transition table and medicine bottle tasks but otherwise we will stick to positioning and practice.

We now have 3 tasks programmed and two with an almost complete programme that needs a little tweaking. Overall the whole thing feels a lot better.

The project is almost complete. we will be meeting with our design mentor next Wednesday to make a prototype of our product and then the script can be completed and rehearsed. The team need to perform for an audience as part of the process so we need to get cracking with the acting parts and learning the lines. Then it's practice, practice, practice!!!

Time to do things well

Our unexpected extension has had the lovely side effect of being able to do things well. We can take the time to tweak the project and really try to get the robot to work as well as we can.

For the project we asked another mentor, Sophie, to help us with the design of our solution. Sophie is a prop/theatre designer by trade and we asked her to apply her skills in tweaking our product design. We had reached a stage where we had two alternative solutions to our problem and the team were finding it difficult to move forward and choose one. Sophie parachuted in and applied fantastic design principles to our products. She took the team back through the process, the decisions they had made so far and why they had made them. The team then re-checked the products available in the market and considered who might actually buy the products they had come up with.

The team have made a decision about which product they want to go with and the next stage is to make a prototype, this is quite tricky as the chosen material is most likely to be plastic - we shall wait to see what ingenious solutions they came up with to solve that one.

We are making slow but steady progress with the robot game. The robot can now reliably complete two tasks and hopefully these will be joined by a few more over the next few weeks. The first task has shown us the absolute importance of positioning the robot accurately and so this gives us a heads up on the programming. choose EXACTLY where the robot is going to start from and design your programme from there, measuring carefully each time (with a tape measure if necessary) before we set it off.

The downside of the extension is that it feels like a long slog now and some of the team are getting quite tired. I hope today's success and the possibility that we now know how to tackle this thing will keep us all going just a little bit longer.