Radioactive Roomba
Team Members Heading link
- Erich Christ
- Keefe Li
- Brian Mlot
- Ramsha Shahid
- Joanna Tan
Advisor: Jonathan Komperda
Sponsor: Fermi National Accelerator Laboratory
Project Description Heading link
The goal of this project is the development of a robotic cleaning mechanism (RCM) that collects radioactive dust off the floors of the particle accelerator tunnels located at Fermi National Accelerator Laboratory (Fermilab). When the particle accelerator is active, it produces residual radiation that is absorbed by dust in the tunnels becoming a radioactive hazard. After researching dust collection options such as vacuum, tape, and electrostatic surface, it was determined that microfiber cloth was the ideal choice. The final solution is a cleaning mechanism that continuously feeds in new material of microfiber cloth that is pressed against the ground to wipe and absorb dust particles. The cloth is then rolled up and collected for easy disposal. An initial experiment showed that microfiber cloth has a dust capture rate of 85.6% per 5 grams of dust in 9.64 ft^2 per 1 ft^2 of cloth and is considered the optimal performance metric. A second experiment using the prototype RCM showed a dust capture rate of 39.0% per 5 grams of dust in 2.95 ft^2 per 1 ft^2 of cloth. Utilizing the second experiment’s data and extrapolating the cost of building and operating the RCM to the 94,000 ft^2 facility at Fermilab, the potential five-year savings plan for Fermilab versus their current outsourcing cost is $298,365. Future iterations would focus on improving the dust capture rate to match that of the first experiment. Such iterations would include adding a dynamically actuating mechanism to the system that presses the cloth to the ground. The refined system would form to the deviations in the floors thereby allowing for a higher dust collection capacity from the RCM. The future end goal of this project is to create a fleet of autonomous RCMs that would clean the tunnels periodically to reduce buildup and overall operation time.
See supporting documentation in the team’s Box drive.