We held our first team meeting in the local library today where we decided on our strategy for PiWars. We created a project document to track requirements and project risks and record progress and shared it on google docs and decided on the following five guiding principles:
Prioritise the challenges
Remote controlled challenges are to be targetted first, followed by autonomous challenges.
Vision challenges to be kept until last. We know a student at the Edinburgh University who might be interested to jump aboard. We could task him with the Hubble Telescope Nebula Challenge. Arul has been tasked with persuading him aboard - fingers crossed.
Build a drivable robot platform ASAP
This will enable the kids on the team to practice driving the robot and learn how it responds to joystick controls. One of Brian Corteil's oft repeated pieces of advice is to get lots of driving practice. It's repeated THREE times in his list of hints (rules 4, 8 & 18).
Build TWO platforms
Our team is split across two families. Two platforms allows both sets of kids to practice driving and both groups of team members to develop our solution in tandem. The second platform also acts as a backup strategy on competition day should our main platform suffer a brain-fart.
Target completion rather than points
As this is our first time competing at PiWars, it's more important for us to understand and develop the basic solutions to the autonomous challenges and not get side-tracked with clever potential optimisations for extra points. Once a challenge is completable, the payback for improvements is small compared to completing another challenge.
As soon as we have a solution that works (any solution... no matter how slow) we should move on to the next challenge. This is known as the MVP principle (Minimum Viable Product) - just enough to get by and no more.
If we have time at the end we can choose to improve on one or more of our solutions. If not, we can always target improvement in our second year. This leads us to our next strategy...
Target camera-free solutions first
The Straightish Line Speed Test (Blast-Off) can be aided by a camera trained on the road ahead to warn you of upcoming turns. However, it is completable with only line following sensors. Rather than burn additional time trying to get that feature built in to our solution we would be better off working towards completion of another challenge.
Also, if we have to drop anything due to time constraints, we should drop the Hubble Telescope Nebula Challenge.