Government agency Innovate UK has channelled £300,000 of funding into trials of an artificial intelligence (AI)-enabled system designed to automate PPE checks in the workplace.
The grant will be used for a real-time video analysis platform being developed by British startup Cortexica Vision Systems and IT giant Cisco.
The trial system will monitor workers as they enter and leave businesses, using live footage from cameras mounted above entrances and exits.
Machine learning algorithms automatically check employees’ PPE equipment, such as headgear, eyewear and footwear, against client inventories. An alert will sound if any workers are flagged as non-compliant and they will be either restricted or denied access.
The system, called AI-Safe (Autonomous Intelligent System for Assuring Safe Working Environments), is designed to streamline PPE checks, typically done manually, and reduce the likelihood of human error.
Vision AI, the technology that underpins the system, developed by Cortexica, is ready and now the team will test it in reality, said Alastair Harvey, chief solutions officer at Cortexica: “The next stage is to apply it in real world environments with all possible PPE combinations, whether this be in tumultuous weather, low lighting or in sterile laboratories.
“We will also finalise how it can run ‘on device’, for example on cameras or computers, without an internet/cloud connection. We are working on pilots with businesses to develop these custom requirements.”
According to information on Cortexica’s website, Vision AI replicates the human ability to locate visually exact or similar matches to an image from a dataset of other images. When used for video analysis it provides “insights into continuous streams of vision”, whether pre-recorded or in real time, with no manual tagging or manipulation.
Harvey added: “We automate video monitoring to check for non-compliant clothing by training the system to identify correct and incorrect PPE and how it is worn. We do this by localising, segmenting and then tracking each human action and its garments.
“This can be done both at the point of entry to the workplace, for compliance, and through continuous monitoring when in controlled space. By automating safety at scale, via unlimited cameras and alert systems, human resources are freed up.”
The firm’s AI techniques were initially pioneered by founders Dr Anil Anthony Bharath and Dr Jeffrey Ng in the labs of Imperial College London, in 2005. The firm now employs a team of 35 R&D scientists, including 15 PhD graduates, who continue to develop and enhance the algorithms and supporting architecture.
AI-Safe is being funded under the government’s Industrial Strategy Challenge Fund, set up last year as part of a £4.7bn boost to research and development spending for business over the next four years.
Dr Nikos Pronios, innovation lead for robotics and autonomous systems at Innovate UK, said: “Projects such as this demonstrate how the development of AI technologies have huge potential to improve workplace safety.
“There is huge commercial potential from the commercialisation of these technologies which is why Innovate UK funds competitions in this area to help UK companies maximise their opportunities.”
We automate video monitoring to check for non-compliant clothing by training the system to identify correct and incorrect PPE and how it is worn. We do this by localising, segmenting and then tracking each human action and its garments.– Alastair Harvey, Cortexica