Ideas for TRAILS in manufacturing
Record how people move and work on the factory floor to rethink staffing and training
In manufacturing, how workers move has a big effect on productivity. On lines that run 24 hours a day, workloads shift with the time of day, and the way work is done varies from person to person. Yet where and how workers move is rarely recorded, so it's hard to know where to improve. A record of routes alone doesn't explain why someone's work stopped at a certain spot.
Common challenges on site:
- Uneven staffing across processes: Some processes have workers standing by while others are short-handed. Shift work and workloads that change with the time of day make staffing even harder.
- Slow first response to problems: When a quality issue or equipment fault occurs, it takes time to find out where the workers who can handle it are. Recording and reporting what happened also takes effort.
- Expert skills that are hard to pass on: Experienced workers' preparation and techniques rarely make it into manuals, and are lost when they retire.
- Hard to see what multi-skilled workers contribute: Knowing who is working where depends on managers' observation. It is hard to objectively capture the contribution of workers who cover several processes, or how skilled they are at each one.
Ideas for TRAILS × the Astra spatial AI camera
TRAILS needs no installed equipment. Pair it with Astra, a wearable spatial AI camera, and you can record each worker's location and first-person video at the same time, even while the factory is running. You keep not just where people were but also what they saw and did there. Here are some ways it could be used.
Use location records to see how many people were at each process at each time of day. For processes with a lot of waiting, the video from those moments shows the reasons, such as waiting for parts or changeovers. This could be used as input for rethinking shifts and staffing.
When a fault occurs, location records can help find which workers are nearby. Based on video and audio from the response, AI could also organize what happened and how it was handled, and draft a report.
Record experienced workers' tasks with location and first-person video. Because the record shows which equipment they visit in what order and what they check along the way, it could be used to pass on skills and create training materials.
Use location records to see, by person and by process, how much work was done at each process. With Astra's 3D hand pose estimation capturing hand movements during work, you could also compare steps and movements with those of experienced workers. This could help you understand the range of processes each person covers and their skill at each one, as input for fair evaluation and for deciding which processes and skills to learn next.
* The Astra spatial AI camera is currently in development. This page, including 3D hand pose estimation, shows possible uses, and the actual product specifications may differ.