Innovation in Motion: FDOT’s Approach for Artificial Intelligence Success
With a clear policy, pragmatic guardrails and pilot-first execution, FDOT is turning AI into measurable operational value — while keeping people at the center of every decision
Rudy Powell, PE, Chief Engineer of Operations | Florida Department of Transportation
Artificial Intelligence is resetting the baseline for how people work and how organizations operate. It is a true game-changer for transportation agencies. I equate AI to the revolutionary technologies that replaced our drafting tables with computer-aided design and that replaced massive mainframes with desktop computers. At the Florida Department of Transportation, AI is already elevating accuracy, accelerating delivery and stitching together data and processes across the enterprise — and we’re just getting started.
For decades, FDOT has treated technology as a core capability — not a side project — from enhanced operations to data‑driven decision-making. Our move into AI follows that pattern: build the right guardrails, invite rapid experimentation and scale what works. In May 2024, we published FDOT’s first Artificial Intelligence Policy, one of the earliest among state DOTs. It sets clear standards for human oversight, transparency, privacy and security — so our teams and partners can innovate confidently in the years ahead.
These include:
- Keeping humans in the loop for all AI-influenced activities
- Being transparent about AI’s role in our work products
- Complying with all laws and rules
- Protecting privacy of individuals and the department
- Ensuring security of FDOT’s AI tools and access to them
We crafted our policy to highlight critically important components while setting the guardrails for execution wide enough to allow innovation, creativity and learning to occur as we move forward.
A Template for Integration
Building upon the policy is our AI Technology Implementation Framework, which provides teams with practical guidance on identifying use cases, ensuring secure applications and practicing responsible innovation. Complementing this guidance is a list of AI tools that our department has approved for use after assessing various options in the marketplace.
Think sprints, not marathons. FDOT’s AI strategy favors pilots over giant leaps because speed and safety can coexist. Pilots keep risk low, costs contained and momentum high. We focus on learning. What matters is the learning curve, and every turn is guided by our AI policy.
Our top priority is always to accomplish our mission: to ensure a safe, efficient transportation system. As we seek to make this system better, faster and smarter through AI use cases, we are using a five-step process:
Step 1 is keeping our AI Policy front and center.
Step 2 is generating use cases for AI integration and filtering them to find cases that have the highest ROI.
Step 3 is drilling into the feasibility of these identified use cases by asking questions like: Can we get the data that’s needed? Do the necessary tools exist? How will we address risks relating to privacy, security and transparency?
Step 4 is determining the value proposition — the tangible and measurable outcomes we expect to achieve in a use case. These might involve time, dollars, quantity, quality, speed or other key performance indicators. Importantly, we want to invest in cases that can be scaled up and replicated in other parts of the department.
Step 5 is sustaining a vibrant feedback loop to ensure that the team’s experiences, lessons learned and new ideas are circulated through our collaborative platforms.
These steps will serve us well as we encounter emerging opportunities for AI use cases while refining our existing deployments.
AI Policy Highlights
- Human in the Loop – AI must supplement or complement work accomplished by humans and must engage humans throughout the process; Humans are fully responsible for the work and product outputs involving AI.
- Transparency and Accountability – We must disclose if the products are generated partially or fully with AI assistance.
- Ethics – We must comply with all applicable laws, rules, regulations and policies.
- Privacy and Data Protection – We must protect people’s privacy and comply with all applicable data protection regulations, ensure human validation of the AI data and output, and protect information that is exempt from public disclosure per Florida’s public records laws.
- Security – Employees, vendors, consultants and contractors are prohibited from attempting to gain access to AI applications not approved by the Florida DOT.
Inspiring Collaboration and Innovation
Workforce readiness is a key aspect of our approach to AI. We see AI as a workforce augmentation strategy. Our focus is on enhancing, not replacing, the talent in our organization. Accordingly, we have endeavored to engage our team “where they are” in their understanding of AI. We anticipated that some, the early adopters, would start asking questions and generating ideas for use cases, even though others might adopt a wait-and-see approach. Still others might simply be change-averse, which is common when once-in-a-generation technologies like AI come along.
With these realities in mind, we built our AI Hub, a digital platform that serves as the central point for “all things AI” within the department. First and foremost, the Hub is an educational resource that lets people learn about AI, explore various tools and see the ways our agency is employing them. The Hub also lets individuals submit ideas for using AI to yield greater benefits and lets them request the relevant software. Over time, we plan to evolve the Hub by adding new information and resources, as well as new AI tools, all while retaining alignment with our policy.
We are eager to engage early AI adopters to capture their energy and innovation. As such, we created an offshoot of the AI Hub, called Community of Practice groups, which support collaboration within the Microsoft Teams environment. An organizer leads each group, and people sign up to take part in a true back-and-forth dialogue to share experiences and latest takes on various projects. Participants also can test drive AI tools to see what might work for their use cases.
For decades, FDOT has treated technology as a core capability — not a side project — from enhanced operations to data‑driven decision-making. Our move into AI follows that pattern: build the right guardrails, invite rapid experimentation and scale what works.
– Rudy Powell, PE
Chief Engineer of Operations, FDOT
AI’s Promise – 3 Areas
As AI matures and adoption accelerates, three areas stand out as being transformed by the use of AI. They share a common challenge: processing volumes of data that defy human scale and converting it into decisions in near real time. Those areas are:
1.) Achieving Broad Operational Improvements
In the near term, we anticipate improvements to traffic operations by using AI to impact incident response, traffic monitoring, signal timing and wrong-way driving. In the longer term, we intend to tap AI’s power for traffic prediction and congestion relief.
We see the sharpest near‑term gains in emergency response. In evacuations and storms, our teams process more data than any person can absorb; AI helps us separate signal from noise and act with confidence. We can detect flooding, assess asset conditions, refine detours in real time and choreograph recovery efforts. And we don’t stop there — the data we gather during response cycles back into our inspection and work programming, making the network more resilient with every event.
2.) Integrating Connected & Autonomous Vehicles and Advanced Air Mobility Technology
Most new road vehicles purchased today have some level of autonomous functioning or safety features that are poised to take ever-fuller advantage of AI. At the outset, these features can improve road safety by letting us get real-time information and alerts to vehicles and drivers. Additionally, as vehicles communicate more fluidly with infrastructure, it completes a circuit of communications that can further enhance safety and system performance. Vital information will reach first responders faster, and construction and maintenance workers in work zones will get heads-up as dangerous scenarios unfold. In short, the integration of vehicles and infrastructure, aided by AI, will help us and system users make decisions faster, and in turn will improve safety and make the whole system more efficient.
We also have our eye on the emerging interstates in the sky. FDOT is a leader in embracing Advanced Air Mobility (AAM), which encompasses a new generation of aircraft types, including vertical takeoff and landing (VTOL) aircraft, that are poised to serve people’s intracity and intercity travel needs. Florida was the first state to formally launch an AAM strategy, and we intend to remain a leader in this area. We will integrate these revolutionary vehicles and related facilities into our overall transportation system to give Floridians access to the fastest, most cost-effective travel options modern technology can offer.
3.) Simulation, Modeling and Predictive Analysis
Many new opportunities are arising as we step beyond 2D plans to create 3D digital twins of our built or planned infrastructure. This will let us see the impact of infrastructure changes on road users and all modes of transportation. For example, using a digital twin model, we can see how a new intersection or interchange is going to work with all the predictive analytics that it affords. Our team will have the power to see the impacts in the model environment, ahead of time and in real time, and integrate that information into the design and development process.
Strong, continuous collaboration is key — particularly when the technology is constantly changing. By establishing clear AI policies and processes, focusing on step-by-step pilot projects and investing in sophisticated collaborative and educational tools, we are confident that our team will successfully harness AI to make FDOT even more effective in achieving its mission.
ABOUT THE AUTHOR
Rudy Powell, PE
Chief Engineer of Operations
Florida Department of Transportation
Rudy Powell, PE, is the Chief Engineer of Operations for the Florida Department of Transportation (FDOT). He oversees the department’s work in the areas of safety, emergency management, research and innovation, traffic engineering and operations, forecasting and performance, and maintenance.
His previous roles at FDOT, which span 20 years, include Director of the Traffic Engineering and Operations Office, Interim Director of the State Materials Office, Director of the Office of Maintenance, State Construction Engineer, State Specifications Engineer and Structures Area Engineer. Earlier in his transportation career, he spent more than a decade in private-sector engineering roles.
Powell earned his bachelor’s degree in civil engineering from The Citadel Military College of South Carolina and is a licensed Professional Engineer in Florida and South Carolina.