the promises of artificial intelligence to improve prevention


The SCAR robot dog, developed by the Ecole des Mines de Nancy, in the underground laboratory of Bure (Meuse), May 11, 2021.

When accidents occur in a company, the reports are often written spontaneously, without formatting. Free words that provide many details, a rich context, but have their downside: they are very difficult to exploit. A problem that artificial intelligence (AI) tools could overcome, with their ability to bring out recurrences of weak signals from a textual mass.

Can AI be an asset for the health and safety of workers? This is what many scientists think today, observing the growing potential of technological solutions and the promising prospects they open up for prevention in the workplace. “The progress is impressivenotes Martin Bieri, in charge of studies in the digital innovation laboratory of the National Commission for Computing and Liberties (CNIL). In organizations with high-risk occupations, where a lot of accidents are observed, AI can be a valuable aid in exploiting data and better understanding the multiplication of incidents. »

Solutions are already deployed in some of them, such as the Plus software from the company Safety Data – Omnicontact. He must “improve security through the automatic processing of textual data and act as a decision support”, explains one of its employees, Céline Raynal. Other tools aim to monitor workspaces and their occupants in real time, with the sole declared aim of securing them, and to issue an alert if a danger is spotted.

Read also Article reserved for our subscribers “The development of an ethical artificial intelligence is a key issue for companies”

Some are already deployed in organisations, such as cameras on board Blaxtair construction vehicles, which detect pedestrians using a person recognition system. Others are still being tested, but could find their way into enterprises in the coming decade.

The help of collaborative robots

This is the case of connected equipment capable of measuring and transmitting employee biometric data (heart rate, for example) and joint positions. “The AI ​​will be able to give the alert on the fatigue of a worker by combining the analysis of different parametersexplains Timothée Silvestre, in charge of forecasting at the French Alternative Energies and Atomic Energy Commission (CEA). The recording of certain movements or the accumulation of weight carried may also be carried out with worn sensors, in order to prevent risks for the employee’s body, in particular musculoskeletal disorders. »

You have 57.53% of this article left to read. The following is for subscribers only.



Source link