Safety App Streamlines Data Analysis, Pre-Task Planning

Fiercely protecting the safety of all is a core value at TDIndustries and a priority at every level. TD has heavily invested in advancing training and safety technologies throughout the organization, staying at the forefront of construction safety innovation for many years.

Since 2014, the team has won national trade association and other industry safety awards each year. Vice President of Safety, Health and Environmental Jamie Dabbs, CSP, recently answered a few questions about how the TD safety team has been using AI for better planning and results.

Q: When and why did you start using AI as a tool to improve safety for TD and its customers?

A: We started using AI to evaluate, score and grade our daily planning conversations in early 2020. At TD, we complete more than 400,000 safety processes every year, and with over 25 safety professionals on any given day, we can only effectively evaluate all plans if we use technology.

In safety, we use a DMAIC process – define, measure, analyze, implement and control. We needed an AI system that can take a large amount of unstructured human data and use it to make structured reports that we can use strategically to improve. Every year we evaluate thousands of daily planning conversations and score hundreds of thousands of pre-task safety plans using AI.

We piloted an integrated AI assist feature into our pre-task safety plans this summer. We add detailed tasks to the safety app, and it creates a detailed work plan with a job hazard analysis within seconds. We will always require our experience and critical thinking to assess the project plan, but the AI assist reduces the time spent writing and can fill in gaps that may have been overlooked.

Q: How are you planning to use AI next?

A: Next year, we will implement AI into our safety observation processes. For example, a user observing a crane installing a rooftop unit can input the description of what is happening. AI can then share (1) a list of safety considerations to monitor and (2) a list of safety-related questions to ask the operators performing the tasks.

Q: How does AI benefit TD's customers and employee-owners?

A: One of the most evident benefits is that it reduces the time needed to build a safety plan. It also improves the flow of communication between field leaders and crew in the daily planning conversation. Based on the detailed tasks entered by the leader, the AI can outline the discussion with a structure that can make it easier for crew members to communicate and follow.

Instead of employee-owners having to create a pre-task plan manually on the fly, it allows safety plans to be enhanced from our learnings across the company and our subject matter experts. When a team member manually adjusts an AI-generated safety plan, the model learns to add that piece the next time someone requests a similar plan. This allows us to improve our safety processes more quickly and collectively.

Beyond playing a role in improving quality, flow and efficiency, the AI model also dramatically increases the user experience and design and can easily translate into Spanish or other languages.

Q: Do you anticipate additional uses for AI in construction safety?

A: Yes, AI is developing quickly and will inform best-in-class safety processes in the near future. It will use our learning points to prevent issues we’ve identified from recurring.

It will also get better at cross-zone resourcing. Currently, AI is a language model that focuses on one system, like video recordings or text-based plans. But in the future, it could pull from multiple sources like lessons learned, our TD safety manual and more. Pulling and checking across these multiple systems will allow the daily planning conversation to get beyond safety and integrate quality, lessons learned and other aspects of the work.

Q: How has TD's innovative use of AI been received?

A: The safety team has quickly adopted the daily use of AI in the app. We have focused for years on continually improving our daily planning conversations and have worked toward using a job hazard analysis library as a foundation for everyone’s daily plan. Our use of AI has allowed this goal to become a reality and has made safety plan creation and evaluation more efficient. To highlight our success, TD, along with our industry partner, presented on this topic at the National AGC Safety & Health Conference in Nashville and the TEXO Safety Professional Day Conference in Hurst, Texas, this year, and we had the opportunity to highlight our innovations in this area.

It is exciting how quickly AI is becoming a resource multiplier, filling knowledge gaps and improving the quality of construction safety observations, daily  plans and conversations.