Work to Zero, Safetytech Accelerator Share Results of Pilot Program to Advance Workplace Safety

Our mission at the Work to Zero initiative, a National Safety Council program funded by the McElhattan Foundation, is to eliminate workplace fatalities through the use of technology. As innovation continues to advance year after year, safety progress is becoming even more achievable. To help us in this pursuit, Work to Zero partnered with Safetytech Accelerator in 2021 to enable small- to medium-sized companies to test and adopt safety technologies within their various industries. Recently, we concluded our first pilot program together, which yielded some noteworthy results in this quest to mitigate serious injuries on the job. 

Pilot Background

As part of the collaboration, Work to Zero and Safetytech Accelerator hosted their first-ever Safety Innovation Challenge in 2022, which focused on identifying solutions to prevent injuries associated with working at height, a leading cause of serious injuries and fatalities in the workplace. In fact, in 2021 alone, falls from height contributed to 645 workplace deaths, about 14% of total workplace fatalities, according to the Bureau of Labor Statistics. During the challenge, six companies showcased their technology solutions, including robotics, drones, computer vision and video analytics.

However, we didn’t want this important work to stop after the challenge, so Work to Zero and Safetytech Accelerator introduced a pilot program where one of the competing companies would partner with another organization to further explore how technology can prevent injuries when working at height. Ultimately, Service Center Metals, an aluminum extrusion and billet company, agreed to participate in the pilot. They selected Matroid, which specializes in software for building, deploying and managing computer vision detectors, to trial their technology. 

Pilot Results

From January to May 2023, Service Center Metals utilized Matroid’s computer vision technology within its shipping and receiving docks to detect damaged, missing or under-inflated safety airbags that help break potential falls during the loading and unloading process. Since workers climb on and off trucks up to 14 times per hour on average, these airbags are crucial to preventing workplace falls and injuries. Even though workers undergo regular safety training, more needed to be done to monitor the airbags and ensure everyone’s safety. That’s where Matroid’s computer vision technology made a significant difference for Service Center Metals, including:

• In the two months following deployment of the technology into the loading bays, Service Center Metals reported zero near misses and injuries requiring first aid
• A substantial increase in daily airbag compliance rates from 25% to more than 90%
• A 400% increase in unsafe observations by workers because of the renewed focus on safety through the pilot program, allowing the team to dedicate more time to other safety challenges
• Improving the company’s digital readiness from the adopting to the integrating phase, with evidence of overall improvements in safety compliance and workplace culture

Key Industry Learnings

The pilot program also presented several key findings that can help all employers as they look to implement technology-driven safety solutions in the workplace, including: 

• Computer vision may be a viable tool for the early detection of critical safety risks
• Piloting technology necessitates comprehensive planning and should be approached as an ongoing, iterative process of continuous improvement
• Having a formal and informal communications strategy in place with employees, while operating with transparency and openness, can help streamline the implementation process for tech-related safety measures
• Barriers to adoption may include resource constraints, employee buy-in and assessing the return on investment

To learn more about this pilot program, check out the video below, or read the full report: Safety Innovation Challenge: Mitigating Work at Height Risks Through Computer Vision Technology.