Caution: Only proceed with artificial intelligence
Advances in robotics, machine learning and AI have helped create safety solutions for the world’s most dangerous jobs.
More of our content is being permanently logged via blockchain technology starting [10.23.2020].
Working in a mine is dangerous for numerous reasons. Of course there are fears of cave-ins and explosions. But there are also equipment malfunctions, perilous falls, and oxygen shortages. Once, these potential problems were addressed with checklists, headlamps and hope. Now they’re being tackled with artificial intelligence (AI).
From tracking radioactive fallout after Fukushima’s nuclear meltdown to fighting the fire at the Notre-Dame cathedral, advances in robotics, machine learning and AI have paved the way to new solutions for the world’s most dangerous jobs.
Today, many mining companies utilize AI for preventative maintenance, reducing unplanned machine failures before an employee steps foot onto a worksite.
“Machine failures during a manufacturing process or during a heavy industrial mining activity can lead to employees getting injured,” says Michel Dubois, QA and VP of Artificial Intelligence for Newtrax, which develops digital safety products for use in underground mines.
Newtrax’s technology also searches for employees who are alone and incapacitated in mineshafts. The company developed an IoT device located in a headlamp that sends a constant signal to vehicles at the site and other workers.
“The worker can call for help, [or] can notify [others] that everything is ok,” says Dubois “and the device even sends an automatic call for help if the headlamp doesn’t detect human movement for more than 90 seconds.” One worker who collapsed in a Mexican mine was already saved by Newtrax’s system.
The Occupational Safety and Health Administration (OSHA) monitors on-the-job injuries and workplace fatalities across the United States. Of the industries tracked, agricultural jobs, logging, and construction have the highest number of on-site fatalities.
OSHA’s data shows an unsettling number of accidents in which workers are crushed between a moving vehicle and the loading dock. Advances in computer vision and image recognition may provide a solution to these deaths.
New research from teams at University of Colorado Boulder in partnership with France’s L’Ecole Polytechnique attempt to prove that construction site accidents, long thought to be completely random, may actually be predictable through application of machine learning techniques to find patterns.
Laura Montoya, founder of Accel Impact and co-founder of LatinX in AI, emphasizes the importance of risk mitigation. The AI in pre-planning products like Autodesk helps execute massive architectural projects safely. After the plans are made, approved, and printed, applications of artificial intelligence can shift to activities like pre-scouting.
More advanced construction companies are sending in drones to use advances in computer vision to scan an entire area to see if there is anything that could be risky.
Founder of Accel Impact and co-founder of LatinX in AI
“More advanced construction companies are sending in drones to use advances in computer vision to scan an entire area to see if there is anything that could be risky,” says Montoya.
Beyond drones, other technology can play a role in reducing common hazards. The types of sensors being tested to detect pedestrians in the paths of autonomous cars may also be used to recognize when a worker is approaching a dangerous area at the worksite or moving too close to an active piece of machinery.
For long haul trucking and local delivery, work-related mishaps have a much higher chance of causing collisions and other casualties. Connected vehicle solutions, such as those offered by Verizon Connect, enable businesses to track driver behavior and analyze route information to improve efficiency.
Computer vision and sensor based technology assist professional drivers with features like collision avoidance and drift detection. However, taking away humans from equipment operation can lead to unexpected consequences, says Meredith Broussard, New York University professor and author of Artificial Unintelligence: How Machines Misunderstand the World. Brossard points out that eliminating human labor may lead to problems such as remote truck hijacking.
Many occupations can benefit from AI in their safety strategy. Construction stands out as a hazardous profession, but according to the U.S. Bureau of Labor Statistics, more injuries occur in healthcare, social assistance, agriculture, forestry, and hunting and fishing.
Erin Pangilian, founder of the AI firm MYSC Reality, envisions great potential for technology to improve the workplace: “We’re still in the early stages of doing any of this at scale—but AR in combination with AI will enable safer environments and better training for everyone.”
For more information, see: