Wearables transforming safety management in high-risk industries

But worker privacy must be considered

Wearables transforming safety management in high-risk industries

In a recent survey of about 1,000 safety professionals, more than one-half of respondents said they favoured using wearables to track safety risk factors. Professionals most in favour, not surprisingly, tended to work in high-risk industries: manufacturing, construction and oil, energy or gas. At the same time, the survey, conducted by the American Society of Safety Professionals (ASSP) Foundation, also revealed respondents had many concerns about the use of wearable sensors in the workplace.

In recent years, advances in wearable technology have taken safety management in new directions. With sensors or wearables embedded in personal protective equipment (PPE), safety managers can gain a great deal of information, from the number of times a worker lifts a heavy object correctly to that worker’s level of engagement in safety. A variety of new systems have recently come on the market, and while they should not be seen as a silver bullet, wearable devices offer innovative ways to improve safety in industrial workplaces.

Sensors allow the safety manager to collect data on a wide range of factors. Wearables can be used to measure motion in specific parts of the body, which helps in determining whether any particular worker is properly performing a task, such as lifting, or whether a task is badly designed. They collect data about the environment: temperature, noise level and hazardous atmosphere. Wearable technology can also detect the presence of physical objects in the workplace that often cause “struck-by” injuries.

A major purpose of wearable technology is to notify workers of problems so they can modify what they’re doing. Their two-way communication capability and location detection is useful in any industry that employs remote workers. Industries, such as mining, where fatigue is an  issue, have also found them to be useful.



Detroit-based Guardhat combines the traditional hard hat with an Internet-of-Things (IoT) component. The Guardhat system actively monitors a worker’s location, work environment, and even pulse and body temperature. The proprietary software platform collects and analyzes on-the-job data and can provide real-time alerts to the user and the supervisor in the event of an incident, such as a fall, and of potential workplace hazards, such as exposure to toxic gases, lockout zones and proximity to moving equipment.

“We combine our advanced Internet-related technology with dependable safety equipment to create a 3D space and holistic view of every user’s work environment,” says Anupam Sengupta, co-founder and chief technology officer at Guardhat. “This allows us to actively predict and prevent workplace accidents, which ultimately provides a better way to protect people.”

The key piece of PPE in the Guardhat system is the “smart” helmet. It provides situational awareness to prevent many kinds of injuries. For example, when a moving object, like a heavy construction vehicle, approaches a worker, sensors in the hat “talk to” sensors attached to the vehicle. The hat’s sensors monitor the vehicle’s proximity and, if they find the worker and the moving object are dangerously close, an alert is sent to both the worker and vehicle operator, who can then get out of the way.

In the case of a fall, the hat immediately detects the incident, determines the worker’s location and sends an alert containing the location’s co-ordinates to a safety control centre. At the same time, the hat alerts workers nearby and activates its SOS features. The centre makes video and audio contact, sends help and stays in touch with the worker until help arrives.

With Guardhat, too, a worker’s physical condition is constantly checked and when irregularities like an erratic pulse are detected, an alert is sent to the safety centre. The hat will also detect hazardous environments, such as gas leaks, and dangerously high levels of carbon monoxide, radiation and harmful chemicals. It then alerts both the workers and safety centre.

Another innovative function of the Guardhat safety system, Remote Guidance, provides live assistance to a worker caught in serious, immediate danger. For example, it can guide a miner trapped in a darkened tunnel to find the quickest way out. It allows for video and audio transmission, enabling communication between the worker and the safety control centre. After the worker notifies the centre of the problem, the hat sends video to the centre. Using audio-visual feeds and geo-location, the centre can then provide live guidance to the worker and stays in touch with the worker until the situation is resolved.

The safety software collects data on all the incidents it tracks. That data is used to produce analyses: for instance, to develop ways to increase worker hazard awareness and to identify environments that often negatively affect workers’ health.

Corvex Connected Safety’s open, IoT software platform also uses sensors and wearable technology to help safety managers prevent and predict incidents. This system centres on the “Core,” a personal mobile device that looks like a smartphone and is worn by workers at all times, says Ted Smith, founder and CEO of Corvex Connected Safety, based in at Eden Prairie, Minn. The mobile device is connected to sensors on the worker’s “smart” PPE and to the supervisor’s personal device. The worker can input information about potential hazards in the device and send it to a supervisor in real time. The device also allows the worker to communicate with co-workers and receive notifications and safety information from a supervisor.

Smart sensors embedded in workers’ PPE monitor environmental conditions, such as temperature and noise, and ergonomic hazards, such as repetitive stress and struck-by hazards. These data go to the Core, which will alert the worker and provide safety directives.

The system also allows a company to designate zones within the workplace where access may be controlled or where, due to certain hazards, workers must always be wearing specified PPE. Companies define the zone by placing Bluetooth beacons on the area’s borders. Messages from the beacons to a supervisor will let the supervisor know if a worker has entered a zone and is not wearing the required PPE.

Smith says the main purpose of the “connected” system is to improve safety by increasing worker engagement. The system is interactive, and the ability to communicate with supervisors and peers gives them a stronger voice. They are encouraged to make hazard observations and participate in surveys. They know their concerns about potential safety hazards are reaching a supervisor and — because there is a digital, trackable way to measure how long it takes to correct unsafe conditions — they know problems will be fixed.

However, Smith cautions against regarding sensors as a silver bullet and thinking that putting a sensor on a worker will keep them safe. Relying too much on this technology may create complacency about the need for proper training.

For example, a manager may put sensors on PPE to reduce injuries caused by improper lifting. But now the workers may rely on the sensors to tell them if they are lifting improperly.

“For a worker doing some lifts during the day, rather than trying to determine through sensors exactly whether a lift is a good lift or a bad one, our approach is to make them aware that they are lifting a lot and train them to do it the proper way, as opposed to relying on these sensors to say, ‘Hey you’re doing it wrong,’” Smith says. “Moreover, if I am lifting improperly, and for some reason the sensor isn’t working or isn’t in the right place or something else is wrong, I think I’m doing everything fine but I’m not.”



While wearable technology is probably used most commonly as a means of intervention to prevent injuries, it is also being used to produce behavioural models built on the data collected from the dozens of safety-related decisions workers make every day.

The behavioural approach is based on the notion that knowing what to do, and even being aware of imminent risk, does not necessarily determine behaviour. Instead, our actions are often irrational; many psychological, social and emotional factors affect decision-making.

Using the data collected from various sources, including wearable sensors, behavioural analysis seeks to reduce incidents by measuring an individual worker’s risk tolerance level. Safety managers will then understand the worker better and will focus on making the person aware of dangers. They will, for example, be able to send safety reminders to a high risk tolerance worker on the specific risks of a dangerous zone just before the person enters that area.

Jennifer Weeks, associate at BEworks, a Toronto-based management consulting firm, says behavioural data collected from sensors play a key role in allowing safety managers to encourage safe behaviours. It all has to do with providing positive reinforcement.

“The only time workers get any reinforcement of their behaviour is when something bad happens, when they get hurt, for example, because they weren’t wearing their safety gloves. But they get no positive reinforcement for putting on those same gloves every day for years,” she says, adding random, or intermittent, reinforcement is more effective than rewarding every act of good behaviour.

“The opportunity we see with wearables is that if you’re connected to every worker all the time through these wearable Internet-of-Things devices, then you can reinforce their good behaviour on an optimal schedule, such as an intermittent reinforcement schedule. And wearables make it possible to automate this delivery of reinforcement,” Weeks says.

If each worker is wearing the personal mobile device and the PPE is equipped with IoT sensors, safety managers can monitor every time they put on their safety gloves and randomly reward them. So once every four or five times they put on their gloves, they hear a buzz on their device and it says, congratulations, you just received 10 points for putting on your gloves.

“If this happens in a random, unpredictable way, it gets people more excited about putting on their gloves,” Weeks says.

Another major use of sensors is to monitor — and predict — worker fatigue, according to the ASSP Foundation study. Over three-hour increments, the workers who participated in the study wore wrist, hip and ankle sensors while completing common manufacturing tasks. Researchers found that body movement patterns, including walking, change with fatigue. These changes in movement indicate developing fatigue and the need for action, such as scheduled breaks, posture adjustments or vitamin supplements, to relieve the physical fatigue.

The main application of fatigue monitoring is in manufacturing and industries that employ drivers. An Australian company, Smart Cap, uses a wearable band to prevent the micro-sleeps — short episodes of sleep or drowsiness that last anywhere from a fraction of a second up to 30 seconds — that cause accidents. The band can be attached to the inside of a driver’s cap, hard hat or other headwear or be worn on its own.

The band uses EEG (electroencephalography), the brainwave technology used in sleep studies, to determine the driver’s alertness. If the EEG reading indicates a driver is close to falling asleep, the system sends an alert to the person. Sensor data can be collected to produce reports and profiles that will help the driver understand the times of day when their risk is greatest. The technology was first tested in surface mining operations to monitor operators of haul trucks, excavators, dozers, graders and water trucks.

In addition to monitoring fatigue levels, sensor technology is also used to improve driver safety by tracking driving habits, such as speeding and taking corners, that can affect safety.

Designed to prevent improper lifting, New York City-based Kinetic’s wearable device uses sensors and biomechanical analysis to determine whether workers are moving with correct posture. If the device, Reflex, detects excessive bending, twisting or reaching, it vibrates to warn the worker to change his or her stance or get help if the item being lifted is too heavy.

Chicago-based Occly has introduced wearable body cameras that are equipped with a personal alarm, cloud storage, a smartphone app and live video streaming. Alarms are delivered to Occly’s 24-hour emergency monitoring service or to the client’s office. The system is designed to be used in a range of industries, including construction, utilities, education, delivery and retail.



Managers’ ability to use wearable technology to track workers’ activities and location throughout the day gives rise to a concern about worker privacy. The ASSP survey showed privacy was the single greatest concern for the use of wearables at work. As one respondent commented, “I would not want employees to feel their jobs are threatened by not moving correctly or fast enough.”

For Weeks, the issue raises two questions. First, is the worker bothered by the invasion of privacy? Her own research into the psychology behind privacy disclosure, which examined consumers’ attitudes to Google and wearables like Fit Bit, produced a surprising conclusion.

“Even people who report being very skeptical and resistant to revealing personal information will reveal it if there’s added convenience. They may say it bothers them, but it doesn’t actually affect their behaviour, so there’s a limit to how much it actually bothers them. They care more about the benefits of using (Google or Amazon) than they do about the loss of privacy.”

This research seems to indicate that if workers believe the advantages of wearing sensors — for example, the ability to communicate, the element of fun or competition, ease of use — outweigh whatever reservations they may have about privacy, they would not be unduly bothered by the wearable technology.

The second question Weeks has is whether or not there is a fundamental ethical problem with collecting and using this worker data — and the answer may be yes, she says. The rightness depends on how the data is being used. If it is being used to predict risk and make the workers safer, then it’s unlikely anyone would think the use is unethical. The problem arises when the data is being used for a purpose that workers have not approved.

“If the data is sold to advertisers or big-data farming companies, that’s where we start to get into a grey area. But as long as we’re using it to predict safety and identify sites that may need attention, like extra training, then I think that would be fine with the workers. It’s fine in my book,” Weeks says.

With all these options for connected systems, it is important not to overdo it and always keep in mind the problem that needs to be solved, Smith says. There is a lot of wearable technology available, but it may not make sense from a cost-benefit perspective.

“You can put technology in that will tell you if a pair of safety glasses is sitting on a worker’s nose properly, but it increases the cost substantially. You can sensor everything and accept the cost associated with it,” he says. “Or you can take a different approach and try to make workers aware and try to create an environment where if someone doesn’t have the required PPE on, someone else feels free to say, ‘Put your glasses on.’”

Linda Johnson is a Toronto-based freelance journalist who has been writing for COS for eight years.

This article originally appeared in the April/May 2019 issue of COS.