
iStock
COVID has forced organisations to engage people with technology – Elaine Knutt takes a closer look at how tech has accelerated in safety.
The Safety Accelerator scheme from Lloyd’s Register and the Lloyd’s Register Foundation, which matches organisations with a safety 'problem' to tech start-ups offering a digital solution, has brokered 16 pilot projects and recently pushed the first over the line into adoption.
In October, the HSE signed a deal with Safety Accelerator start-up Ohalo, which designed and deployed a tailored data anonymisation solution so that 600,000 RIDDOR reports can be stripped of identifying details before being analysed for insights; Ohalo’s software can deal with the 10,000 returns received each month in just 33 minutes.
Dr Maurizio Pilu, vice president for digital innovation at assurance and certification specialist Lloyd’s Register Group, explains that the Safety Accelerator is targeting the 'safety critical' industries of oil and gas, shipping and food production that Lloyd’s Register serves.
'These sectors are not adopting technology fast enough to reduce accidents. A lot of tech used in other sectors, such as financial compliance, or security, can be repurposed to reduce risk. In some ways, the accelerator is a form of knowledge transfer into the health and safety community.'
Collaboration
Emerging solutions include sensor technology, computer vision systems and data analytics, with the Safety Accelerator working collaboratively with leading shipping operator Scorpio, trialling a system which uses cameras and machine learning to interpret facial expressions and actions – to see whether it can be used to predict accidents.
But Maurizio wants to see more debate around what it calls 'safetytech' – the technology, products and services that enhance safety management in safety-critical industries – with industry leaders, the technology world and regulators openly discussing barriers to adoption, impact and the trade-offs between safety, personal data and privacy.
Maurizio says: 'Privacy and security are top priorities for everyone we speak to. The accelerator can be seen as a 'sandpit' to allow key stakeholders to understand better the safety impact of certain technologies, and better understanding the trade-offs. Solutions are out there that can mitigate or even eliminate privacy concerns, but we just won’t know until we pose the questions in specific use cases. The discussions cannot be held theoretically, it has to be grounded in reality'.
Evolution
Other safety professionals agree that the sector is slow to advance the debate around safety and digital solutions. Adam Wilkinson, quality, health, safety and environment director at CBRE Global Workplace, says that technology adoption has stalled at drones and cloud-based safety management systems.
'More advanced technology tends to get more people focused, then you have a crossover with data privacy, and it starts to get conflicted. That’s when adoption slows down,' says Adam, mentioning systems such as fatigue monitoring or computer vision systems that automate PPE checks.
One problem, he says, is that solutions evolve elsewhere, then entrepreneurs try to sell-in to the safety world. 'Usually the tech comes out of a start-up and is pitched as solving a problem rather than being developed organically inside – it’s not the industry collaborating to come up with tech solutions. It’d be a massive step forward to have problems discussed and formulated and then that goes out to market.'
'I don’t see enough inter-organisational knowledge and learning sharing'
He also suggests that the regulator could help. 'The HSE could give people more comfort in adopting digital solutions, in many cases the guidance hasn’t been updated in years. Even down to things like having a digital permit to work – these are often the last bastion of paper but this is 2020.'
Technology can help safety engagement, he notes, by making processes simpler. For instance, CBRE is in beta mode with a new accident and incident reporting platform that uses 'Natural Language Processing'. If an employee sees an incident or hazard, they no longer have to scroll through a digital form that pre-categorises incidents and then make a subjective decision on which is relevant.
Instead, they write a short report in normal language – like sending a text – and the system makes the categorisation decision. 'To be useful, the data needs to be standardised, so it needs to be automated,' says Adam. 'If you have to scroll through a long list of categories, it sets up resistance.'
Calling for more dialogue in the profession around technology adoption, he says: 'No one will argue that this isn’t the future – we will have to use technology. But the way we get there doesn’t seem particularly joined up at the moment.'
Communication
Construction safety specialist and IOSH council member Jessy Gomes says that the sector struggles to have the kind of inclusive conversations within project teams that are a pre-requisite to technology adoption. 'If people feel incorporated into the conversation, they might be more willing to use new systems,' she notes.
And conversations at company-to-company level are also rare. 'There are bits and bobs of initiatives [on technology innovation] but no synergy, and I don’t see enough inter-organisational knowledge and learning sharing.'
That also applies to transparency over safety data, she says, as a mandate to publish organisation’s performance data – which already exists in some countries, including Canada, and would bring safety into line with pay gap reporting – could stimulate more data analysis and innovative solutions.
Safety professionals with expertise in IT are a rare find; but one that fits that profile is James Sharp GradIOSH, who has taken his health and safety qualifications to NEBOSH diploma level. He was a programmer before joining Riskex to develop its cloud-based safety management system, AssessNET.
'While I was creating the AssessNET platform, I embarked on a professional learning and qualification journey so that I could better understand what health and safety leaders and practitioners need technology to do for them,' he says.
James goes on to explain the value his experience as a safety professional adds to clients: 'The fact that I can understand the pain points, improvement strategies and compliance obligations of our end users has helped me to approach our product roadmap in a client-led way.'
'Automated decision engines need time and data to learn'
He sees the COVID-19 pandemic as an accelerator for technology adoption across enterprises, including the safety function. In the UK’s first lockdown, James and his team created Safe2Day, a COVID health surveillance and tracing platform, which supports businesses to control COVID transmission, address employee anxiety and effectively plan resource requirements by assessing who is and who isn’t fit to work. To deal with privacy concerns, data collected via Safe2Day is held centrally and with GDPR-compliant security.
'The pandemic has forced organisations to engage people with new communication tools, and adopt technologies that enhance connectivity to end users,' he says.
One example of more 'connectivity' that he forecasts is set for growth is 'ring-fencing', where smartphones receive automated safety notifications related to their geographical location: for instance, crossing a threshold into a restricted or mandatory PPE area.
But he predicts that 'machine learning' technologies, which could 'learn' how to predict a problem or risk and 'assess' a situation by analysing multiple data sets – will struggle to gain a hold unless organisations become much more willing to share safety data and collaborate more.
James elaborates: 'Automated decision engines need time and data to learn, so for them to work effectively, collaboration and sharing of safety data between companies will be crucial; GDPR concerns and worries about maintaining control of commercially sensitive information will continue to create blockers that mean the levels of collaboration simply won’t be adequate to drive significant use of decision engines for some time to come.'