Transforming Safety Management at Suffolk with Predictive Analytics

A Safety Model Built In-House, Powered by Posit

By leveraging data, artificial intelligence, and predictive analytics, Suffolk decided to proactively identify where risk exists on their projects and how they can eliminate that risk. To minimize external dependencies and improve reliability, Suffolk brought the safety model in-house, using Posit Workbench and Posit Connect alongside Databricks.

Using Posit Team, the data science team at Suffolk standardized their model workflow and communicate safety scores to stakeholders in two separate ways:

• Safety Dashboard to provide key insights (percentage of high risk projects, any projects with incidents, how predictions compare to historical predictions over time, etc.) at both the project and portfolio level to help project teams see risk in a few different ways
• Weekly Portfolio Risk Email with a concise list of projects flagged as high risk, plus the top features driving each score

The safety model integrates data on staffing, trade partners, observations, incident history, project schedule, and core project details to proactively assess project risk. With the partnership of Databricks and Posit, Suffolk also applied row-level permissions in their interactive applications, ensuring business stakeholders can access a personalized view with only the data they should see through a single, shared app.

Blake estimates that by moving to Posit, the team has reduced deliverability time by approximately 20% simply by not having to search through SharePoint for Excel files.

Fewer Incidents, Faster Insights

"Over the years, we have made significant progress in equipping every employee with real-time predictive insights," said Aleksey Chuprov, VP of Data Analytics and AI at Suffolk. "Those insights power optimal decision-making and best-in-class project outcomes across our seamlessly connected, analytics-driven platform."

Since adopting Posit, Suffolk has achieved a 72% reduction in Total Recordable Incident Rate and a 56% decrease in Lost Time Incident Rate. The team has also cut deliverability time by ~20% by eliminating reliance on scattered Excel files.

Blake highlighted that while technical aspects constitute 80% of data work, the remaining 20% centers on effectively communicating results to ensure end users understand how data can enhance their job efficiency and impact the business. The way data is framed matters too: the initial framing of safety scores as "having a higher likelihood of incident" made it something project teams wanted to avoid, which can affect how people collect data or interpret results. Clearly communicating to stakeholders what the model does and how it can help make them more efficient proved essential.

AI in the Built World

Suffolk aims to be the future leader of AI in the built world, and it is a company-wide initiative to figure out how they can best leverage AI. Some of the initial use cases have included chat agents to retrieve specific information about particular domains and extracting data out of PDFs. Future use cases include augmenting and speeding along the design of buildings themselves and creating entire schedules for project planners. If it currently takes a planner two weeks to complete scheduling planning, the goal is to get them 80% of the way there in a matter of hours or minutes.

Suffolk is also expanding its use of machine learning models into insurance and scheduling, with accompanying Shiny applications and deployed APIs with Git-backed deployment to Posit Connect.

For data scientists interested in taking their first step into AI, Blake recommends checking out the ellmer and chatlas packages from Posit. "It's really easy then to integrate it to any other kind of work that you've done or, if you happen to be using any of the Posit suite of tools, building it into a Shiny app or whatever else. I think that is probably the easiest way to get started."