In high school I did competitive robotics through US FIRST Robotics on Team RUSH. If you aren’t familiar with the program, teams of high school students are mentored by professional engineers to build and compete with a robot. A new game is presented every year in January and the teams have less than 2 months to design and build the robot. This was not ‘battle bots’ – there was a focus on learning and competition.
One of the many lessons I learned on the team was to build duplicates. We had to build and ship the robot by a certain day, but it wasn’t yet time to compete. We would have time at the competition to make mechanical and software improvements to the robot, but we had to have a way of developing them. At the time the Detroit auto industry hadn’t collapsed (the numerous Michigan teams are primarily sponsored by auto engineering firms) so we had enough resources to build our own competition field and and an additional robot. This meant that we could practice our driving and strategy while our competition robot was on the way to the first competition of the season.
Can this lesson apply to physics? The amount of time and money to build the robot seemed immense at the time, but are orders of magnitude less than what it takes to build a particle physics experiment. We certainly can’t build two cryostats out of clean copper and only send one to WIPP. But we did build a mock-up of our time projection chamber to begin the build process. We have such delicate parts that trying the assembly outside of a clean room first meant we had practice before we assembled the real thing. This didn’t even double the cost of the TPC – we had ordered parts with plenty of extras and built the mock-up out of flawed parts that would not have been suitable for the real TPC. Some parts weren’t made of pure materials and others weren’t plated in expensive conducting metals, like the real parts are.
- Fake TPC
- Testing Wires on fake TPC
- Using paper cables on the fake TPC
- Real TPC in the Clean Room
Now we are trying to create a duplicate electronics structure. The detector is built and attached to the read out electronics, so we are debugging the electronics and software to acquire data. While some collaborators are switching board positions and adjusting voltages, I’m trying to fix the software. It is quite the tango – it is hard to tell if some problems are in hardware or software and sometimes one group makes changes that mess up the other group’s debugging. Having a test electronics system with the minimum hardware and computers to test the software means that the two processes will be able to work independently. Once the real electronics system is installed at WIPP (and the bugs are out of the software!) we can use the test system to work on upgrades to the electronics without interfering with the data runs at WIPP. We can even use it for developing prototypes for readout systems for ton-scale EXO.