The question of how best to enhance employee safety on the plant floor has received extra attention in recent months, as manufacturers worldwide assess their ability to protect their workforce from pathogens. Examples include recently published Health Canada’s extensive protocols and guidelines document for plants, the joint OSHA and guidance for manufacturing workers and employers in the United States, or the UK’s official guide to working safely during Covid-19 in factories, plants and warehouses.
The pandemic has highlighted the importance of worker safety in the context of the low-touch economy. Digitalisation, the adoption of which has been accelerated by the realities of social distancing, can play a pivotal role in minimizing the risk of infection for frontline manufacturing employees. As plants reconsider ways of working to minimise contagion, there is a new opportunity to evaluate how digital tools can benefit existing OHS practices and procedures.
More safety progress needs to be made
Although workforce injury rates have fallen in aggregate in the past few years, the declines have not been uniform, and in many cases, not significant. According to U.S. Bureau of Labor Statistics, fatal private manufacturing work injuries fell by 24% between 2003 and 2009. However, the decline was much less impressive over the next 8 years; between 2009 and 2017, fatal injuries declined by a mere 5%.
The issue is not simply that manufacturing safety statistics have largely plateaued. It’s that, compared to other industries, manufacturing remains one of the most dangerous places of work. The incidence of injury in Maintenance, Repair and Operations (MRO) is higher than other work areas. According to RIDDOR (Reporting of Injuries, Diseases and Dangerous Occurrences Regulations), the number of manufacturing workplace fatalities in the UK were 1.5 higher than workers in all other industries in 2018/2019.
Although some of the discrepancy can be attributed to the nature of the work, there is clearly an opportunity and a need for improvement.
How digitalisation can help reduce injury
Digitalisation cannot solve manufacturing safety shortcomings on its own. Digital tools are only as good as the safety practices, processes and people with whom they are integrated. Nevertheless, digitalisation initiatives can make industrial maintenance less risky.
Advanced failure warnings with AI
The ability to schedule and prepare for maintenance tasks can reduce the probability of maintenance workers becoming injured on the job. Unscheduled downtime forces maintenance departments to reassign workers from routine to urgent maintenance tasks to minimize yield loss from the asset, and any other assets that depend on its output, sitting idle.
The pressure to resume production can be dangerous; it is estimated that 60 to 80 percent of all workplace accidents are caused by stress. Working under duress can be further compounded by fatigue if technicians work overtime to resolve this issue.
Moreover, though Reactive Maintenance remains the default in many plants, relying on Reactive Maintenance as the primary maintenance approach is demonstrably an unsafe strategy for maintenance employees. Research from the Belgium Maintenance Association, BEMAS, indicates that when 75% of maintenance is Reactive, the incidence of Maintenance Technician lost-time injuries is 12 times higher than when less than 25% of maintenance is Reactive.
In contrast, AI-driven Industrial Analytics solutions provide early warning of evolving asset failure based on Big Data anomaly detection. Once the Machine Learning algorithms identify abnormal data patterns within the real time Big Data produced by an asset, an alert is sent to technicians warning of impending asset failure.
These early warnings of upcoming failure enable O&M technicians to troubleshoot the issue, consult experts and order parts. Additional information provided by the solution, such as the expected time-to-failure (TTF) and which sensors captured this abnormal asset behaviour help technicians identify Root Cause and establish a repair plan.
By planning for downtime upfront, plants can remove the stress and overtime fatigue associated with unscheduled downtime, turning the unexpected into routine and removing major injury risk factors from repair work.
Servicing and restarting an asset that has unexpectedly shut down can be dangerous, especially in the case of assets that operate at high temperatures or that process volatile chemicals. AI-driven Industrial Analytics’ ability to detect evolving downtime enables technicians to shut down the machine according to reliability guidelines and minimize resulting incidents such as fires or explosions.
To learn how Enlight AI’s AI-driven Industrial Analytics helps cement plants avert unscheduled downtime of high-temperature rotary kilns, click here.
Wearables keep close tabs on workers and hazards
Remote Maintenance began to dominate the reliability conversation because of Covid-19, and the option of going remote has very clear advantages in the changing plant environment. At the same time, certain tasks, such as in-person inspections and bearings mounting, cannot be accomplished from home.
Wearables can provide O&M managers with greater visibility into personnel movements. The pandemic has made this especially useful for enforcing social distancing throughout the plant. At a more basic level, however, wearables can track physiological indicators (pulse, body temperature) to determine whether a worker is in distress, or to warn workers of a hazardous situation in development.
Although wearables must comply with local privacy laws, more precise knowledge about the health and whereabouts of workers while they are in the plant can avert injury and improve emergency response time.
Video conferencing and augmented reality can prevent rookie mistakes
Augmented Reality is also being trialed by manufacturers as a way of providing real-time diagnostic help in the field. Using video chat, maintenance workers can relay images of malfunctioning assets to experts located in remote diagnostic centres. Sometimes dubbed the “virtual shift,” these experts can use augmented reality apps to
- overlay the images with parts or fixes;
- to guide the workers through the remedial process;
- and to prevent unnecessary injuries occurring due to lack of familiarity with the machine.
This type of remote guidance can be particularly helpful for new workers, who are particularly at risk of wrench-time injury. Research by the Toronto-based Institute for Work & Health shows that workers within the first month of working are 3 times more likely to sustain an injury that requires time off. With social distancing measures being implemented across plants and travel limitations, the availability of on-site matter experts cannot be taken for granted.
Deploying drones can remove on-site risk
Drones can be used to inspect asset breakdown, to perform repairs on assets with problems that are of higher risk to repair personnel or assets that are located in areas that are dangerous to access, such as wind turbines in high wind conditions. At present, drones cannot entirely take over in-person asset inspections. While they can see asset issues, they cannot replace the holistic in-person inspection, which in addition to sight includes smelling and listening for abnormal asset sounds.
Nevertheless, although drone usage is not pervasive, they are a viable and affordable method for circumventing maintenance injuries by simply keeping humans out of harm’s way. As Remote Maintenance becomes more and more common, the ability to control drone asset inspections from afar will increasingly benefit plant worker safety.
Summary and conclusion
The acceleration of digitalisation adoption catalysed by Covid-19 can have a significant impact on overall worker safety. Still, it is unlikely that plants will be able to or need to invest in all the above technologies at the same time. In order to prioritise IIoT investments from a safety perspective, plants must evaluate their top risks and hazards. Identifying common and recurring problems can better position plants to assess the safety business case for an IIoT solution.
Industrial plants generate terabytes of process data. SKF Enlight AI is a SaaS Predictive Maintenance solution that uses Automated Machine Learning to identify emerging asset failure patterns within this data. It provides early warnings and sensor-level intelligence to help avert unplanned downtime and meet production goals. For more information on how SKF Enlight AI can improve performance and reliability, click here.