Artificial Intelligence in Occupational Health and Safety

AI Case Study: OHS/EHS Compliance and Hazard Detection

It is hard to equate profitability with spending money on efforts to achieve compliance with Occupational Health and Safety requirements (Environmental Health and Safety - EHS). However failing to adhere to such standards has a direct impact on profitability and the business’s bottom line.

The purpose of this R&D project is an implementation of an AI-based system prototype that is capable of:

  1. monitoring operational compliance to safety requirements specific to workspace where it is installed, and
  2. active anticipation and detection of safety hazards.

The following is a shortened list of some of those EHS/OH&S - related situations that we are currently working on:

  • PPE (hi-visibility clothing, hard hats, masks, googles, ear protection, straps, etc.) use by personnel in defined zones and under specified conditions;
  • vehicle/equipment safe use observations, e.g. speeds/acceleration/deceleration/location/safety zones;
  • bad practices when lifting heavy items by personnel;
  • detection of people with certain characteristics (carrying potential weapons, agitated, demonstrating erratic behaviour);
  • surface monitoring (spilled liquids, shattered glass, nails);
  • people in unusual positions (e.g. detection of incapacitation, falls)

Another purpose of this study is to present an engineering capability to Environmental Health and Safety experts; to demonstrate what is technologically feasible right now; to understand whether any useful outcome can be achieved in this field, what are the constraints and what are the best uses in applying this capability.

Here is a presentation of a part of this project: an AI model that is capable of recognising, identifying and analysing people, safety gear and safety protection zones.

The below video demonstrates how this AI model responds to person movement across several protection zones in construction environment. Safety non-compliance alert is automatically generated when person is detected inside a safety protection zone not wearing safety gear as required for this zone.

Please note:

  1. this setup is a simulated environment filmed outside of construction area
  2. compliance to wearing of Personal Protection Equipment (PPE) is an initial part of this AI study
  3. an ability to detect and classify dangerous conditions as a combination of various factors is the next step that we are currently working on

If your business is located in Melbourne, Australia we would like to offer you a free pilot in using Artificial Intelligence to conduct your OH&S compliance monitoring.
Please do not hesitate to Contact us, we are located in Bayside area.

Skip to relevant parts of the video by selecting below.

0:00 Start of the recording. 'High Visibility' zone is detected. Person is outside this area.
0:02 'Head Protection' zone is detected.
0:03 Person walks into 'High Visiblity' zone. Non-Compliance 'High Visiblity Clothing' Alert is generated.
0:07 Person puts on high-visibility yellow vest.
0:15 High visiblity clothing on person is detected, Non-Compliance alert is cleared.
0:19 'Respiratory Protection' zone is detected.
0:20 Person walks into 'Head Protection' zone. Non-Compliance 'Head Protection' Alert is generated.
0:25 Person puts on red hard hat, Non-Compliance alert is cleared.
0:29 Person walks into 'Respiratory Protection' zone. Non-Compliance 'Respiratory Protection' Alert is generated.
0:40 Person puts on respiratory mask, Non-Compliance alert is cleared.
0:50 End of recording.

This AI case study has been conducted in Australia, Melbourne by FifthOcean and Nola, a foot traffic counter & visitor analytics solution for retailers, venues & events.