Artificial intelligence may report Construction site injuries before they even occur

Artificial intelligence may report Construction site injuries before they even occur

Construction organizations are developing an AI system which predicts worksite accidents --an example of the growing use of surveillance.

A building site is a dangerous place to work, with a fatal accident rate five times greater than any other business.

Now a range of significant construction companies are testing technology which may save lives, and cash, by predicting when accidents will happen.

Suffolk, a construction giant located in Boston, has been developing the machine for more than a year in cooperation with SmartVid, a computer vision business in the same city. Earlier this season, the business persuaded several of its competitors to combine.

Jit Kee Chin, chief data officer and an executive vice president at Suffolk, discussed the project and the collaboration at EmTech Next, a seminar hosted by MIT Technology Review this week.

The machine makes use of a deep-learning algorithm trained on construction site images and injury records. It can be put to work monitoring situations that appear likely to lead to a crash, such as a worker and a building site or even functioning too near a dangerous piece of machinery.

"Safety is a huge issue for construction," said Chin on stage in EmTech. "The typical way safety is managed today is that you attempt to change behavior."

The project demonstrates the capacity to get an AI-enabled personal computer vision to monitor and forecast workplace activity. This is important for the construction industry, which suffers from cost overruns that are severe and bad productivity. Really, the building world has adopted other technology, machine learning, and computer vision.

Suffolk and SmartVid created the Predictive Analytics Strategic Council this March in order for organizations to contribute data which may improve the system's functionality.

Chin says it made sense for competitors to hand over their data because many companies wouldn't have sufficient data by themselves. Deep-learning algorithms require substantial quantities of data to improve their models. Improving safety is an incentive too. "Security was a good place to begin," she said. "Most companies don't have this in-house"

However, while the project is largely designed to improve security for employees, it's also another example of a far wider trend: using AI to track, measure, and maximize work lifetime. Organizations are finding ways to track the work that people are doing and are currently utilizing algorithms to optimize their performance.

That is now a basic part of some jobs, like driving to get a ride-sharing company or working for technology firms like Amazon. And it's unlikely to stop there--we might find ourselves working for algorithms.

Mary Gray, an anthropologist at Microsoft who analyzes the labour behind many tech products, told the EmTech audience a rising number of workers spend their time supporting and responding to algorithms. "It's more than the job we tend to take into account when we speak about automation," Gray explained.