The construction industry, by its nature, can be dangerous. SangHyun Lee, an associate professor in the University of Michigan’s Department of Civil and Environmental Engineering, says wearable sensors can improve construction worker safety and also reduce costs through better data on worker health. He answers questions about his research.
Why attach sensors to construction workers?
Lee: Construction is labor intensive and has the highest number of accidents among all U.S. industries. Many workers wind up leaving construction jobs without enjoying their full career due to injuries or stressful environments.
In addition, construction is struggling to recruit younger workers. It lacks a skilled workforce and that jeopardizes ongoing infrastructure rehabilitation. These wearables can be used to monitor the physical and mental stress of workers so that any potential problems can be understood and, eventually, fixed.
EEG data gives researchers a sense of a worker’s emotional state, and high stress levels can be indicative of unsafe behaviors.
It’s hard for a construction worker to carry any monitoring devices or to stop their labor to provide data. But wearables, like a wristband, can be worn without interfering in that work while continuously sending back data. And with the opportunity for haptic feedback, they can be used for important communication, such as safety alarms.
Beyond the workers, who stands to benefit from the use of sensing technology in the construction environment?
Lee: Some construction companies take steps like having workers stretch before starting the day to reduce the potential for accidents and injuries. They can also provide training to encourage employees to operate safely. So, any improvement for the workers can be directly beneficial to construction companies.
Adding sensing technology could lower insurance premiums since many insurance companies provide discounts if companies engage in activity aimed at reducing workplace injuries and accidents.
Like a patient in health care, we can monitor workers’ physical and mental status to identify red flags. However, this does not mean that all data has to be shared with management, which touches on worker privacy concerns. Workers could be empowered to monitor their own status.
What engineering advances in recent years have made sensing technology more adaptable to the work environment?
Lee: Much of the technology has become truly “wearable” (i.e., comfortable and small-sized) and the sensors, along with advanced signal processing and machine learning, have become capable of operating at a certain standard of accuracy and robustness. Currently, we’re seeing these technologies utilized most in the health care field.
How does your research fit into this overall picture?
Lee: My research applies advanced signal processing and machine learning techniques to derive useful information from wearables. For example, my research group identifies stress, physical demands, risk perception level and other factors from a simple wristband monitor. Sensors track heart rate, skin temperature and electrical activity on the skin.
I work with several construction companies to see how useful these technologies are for worker safety and health.
What are the hurdles to widespread implementation of wearable sensing technology?
Lee: I’ve spoken with many construction companies and their workers, as well as insurance companies. Everyone sees the potential to improve construction safety and health. But they all want to see successful cases of return on investment, which we’re working on.
In terms of the workers, one interesting finding is that older workers would like to see how the technology is useful to them, e.g., actual safety improvements. Younger workers are already familiar with wearable technologies and already see the value. However, they’re willing to wear them if their co-workers are using them, which is more about social influence.
Multiple strategies for the adoption of wearable sensing technology might be useful—targeting different demographics with different approaches.
Source: University of Michigan