Mike Matulik, group leader of the Infrastructure Group in PPD’s Electrical Engineering Department, wrote this week’s column.
|Mike Cherry (left) and Mike Matulik with the robotic system, R2-DO.|
Colliding beam detectors like CDF and DZero are most cost efficient when they’re running and taking data 24 hours a day. Yet these complex devices are prone to occasional component failures. Minimizing these interruptions is a major responsibility of support groups such as the DZero Electrical Support Group in the Particle Physics Division. We closely work with the scientists performing the experiments to optimize the detectors’ performance.
Now we have an extra helper in our group, named R2-D0. This robot, conceived by Mike Cherry, can switch power supplies for DZero’s Silicon Micro Tracker detectors without the Main Control Room shutting down the Tevatron. Other accelerator-based experiments may soon ask for similar robots.
The SMT detectors, installed in the center of the DZero detector, help locate the spot where a proton and antiproton collide. Six crates of electronics, located in the collision hall, collect the data generated by the SMT system, and each crate has its own power supply. These power supplies are heavily taxed. When one of them fails, it can take 10 or more hours to finish the data taking with the Tevatron collider before technicians have the opportunity to enter the collision hall to make the replacement. During this time, the DZero experiment can only record a reduced data set, affecting the quality of the DZero data.
When our group discussed how to reduce the impact of these power supply failures, Cherry suggested—perhaps in jest—that we should install a robot to change the power supplies remotely. When we found no other options that were practical to implement, Cherry proceeded to put together a prototype, held together with C-clamps and operated with a hand drill. He convinced us that the robot could work. We thought “R2-D0” was a natural fit, and we adopted the name without a lengthy Star Wars discussion.
Cherry’s design was refined, and we proceeded to connect all six SMT crates to R2-D0 power transfer switches. Earlier this month, we registered our first R2-D0 save: the power connections from a failed power supply were remotely transferred to a second supply while the Tevatron kept running.
Our group gratefully acknowledges the significant contributions of the DZero Operations Group in PPD to the realization of R2-D0. To paraphrase an old saying: it takes a division to support a great experimental program.