In this paper we address the issue of safety in the use of Personal Protection Equipment (PPE) in construction, industrial, or similar sites where power tools are used. We propose a novel solution that can control actively the power of the tool depending on the worker–tool distance. It is based on RSSI information transmitted by BLE devices arranged in a particular rig, combined with a Bayesian distance estimator. Such an approach minimizes the required instrumentation of the workplace and also the number of configuration parameters; therefore it enables a wide range of applications. Our aim is not only to signal risky situations caused by the misuse of the PPE (either due to its bad fitting or a wrong distance to the tool), but to intervene in a fast and robust way to avoid the safety risk.
This solution is built upon previous results on the statistically sound measurement of distances and closeness in construction sites. Here, we contribute with a thorough analysis of collocating several BLE transmitters near orthogonally, which reduces interferences while avoiding the cost of more advanced technologies. We study how many transmitters are needed and what parameters are the best in the Bayesian filter for the optimal performance of the system.
Real experiments with a prototype have been conducted in a construction workshop where a person operates a miter saw. The results show how the correct use of the PPE (an earmuff equipped with the BLE transmitters) can be inferred from the distance estimation in a robust and reliable way.
