For Shawmut Design and Construction, safety isn’t just a performance metric. Just glance at the company’s core values, and its commitment to safety becomes clear:
“We’re about more than just compliance; we’re about prevention. Our #1 goal is to make sure everyone on site and in the office goes home safely every night.”
Shawmut takes on approximately 500 projects per year. As the company grew to its current size, it needed to continually raise the bar on risk prevention. Shawmut decided to add AI to its risk technology stack to identify, monitor and ultimately reduce critical risk factors.
When tech providers and users work together, everyone wins—which is why our team worked with Shawmut to integrate the new AI and define our shared goals. Below are a few things we learned along the way, which can bring value to any similar partnership.1. Don’t choose goals in a vacuum.
Vinnie, Smartvid’s AI engine, uses construction-specific AI models to identify risk indicators in project data. These capabilities provided context to help Shawmut identify leading risk factors: work at height, housekeeping, standing water, and workers missing personal protective equipment (PPE). Each item is a leading indicator of risk on an active jobsite, and is detectable by Vinnie—a great example of two teams combining their strengths to create safer environments for all involved. Each of these factors is combined into a final risk score for each project - enabling Shawmut to rank projects by potential risk. Interested in how the final project-specific risk score was generated? Shawmut’s VP of Safety, Shaun Carvalho, will explain in detail the specific weights given to each of the leading indicators in this webinar.2. Monitor progress in ways that benefit all partners.
Shawmut’s technology stack includes site documentation from Procore, OxBlue and 360 imagery. Vinnie integrates with each platform, which made it easy to create both an executive dashboard and project-level reports for Shawmut. The dashboard delivered an objective view of risk-related issues across all Shawmut projects, as well as weekly trend data to highlight progress; the project-level reports dove into specifics about key risk categories and shared photo examples. Both help Shawmut synthesize and escalate information as needed.
3. Turn tech insights into human action.
When Vinnie found a high rate of housekeeping observations on a particular project, Shawmut executives took a deeper look at what was going on. The photos Vinnie identified as risky showed that the site had already begun demolition, a high-risk phase of construction that was creating piles of debris and standing water. Focusing attention on the new phase helped manage these risks as demolition progressed further.4. Celebrate wins that matter for both partners.
Shawmut’s “Caught Safe” program celebrates workers who follow safe practices and sites with particularly low risk rates. Thanks to Vinnie’s analysis, Shawmut can identify even more employees in this category. As a result, everyone wins: Vinnie gets to use his AI insights for positive reinforcement while Shawmut incentivizes safe behavior.5. Don’t settle.
A great partnership takes time, communication, trust and effort from all parties involved. It also doesn’t have to end after reaching its initial goals. Shawmut is now part of Smartvid’s Predictive Analytics Strategic Council. This work, and this focus on preventing risks before they occur, ties neatly into Shawmut’s core values, which helped drive the company to seek AI support in the first place.
It’s been a gift to have Shawmut’s perspective on that council, and our collaboration has become a model for AI technology can best work with construction managers and firms.