Abstract Manager

Traffic Operations Centers (TOCs) are increasingly challenged to maintain high levels of safety, reliability, and responsiveness while operating under staffing constraints and growing system complexity. This presentation examines how agentic AI—AI systems capable of autonomously executing multi-step workflows with human oversight—can be pragmatically deployed to automate repetitive, rules-based, and error-prone tasks in live traffic operations environments. Drawing on recent field assessments of TOC standard operating procedures and real-world deployments, the session demonstrates how agentic AI can convert existing SOPAs directly into executable workflows, reducing manual effort while preserving operator control and accountability

Practical use cases include daily roadwork coordination, system QA/QC monitoring, automated incident and vehicle alert processing, and multi-agency notification workflows—tasks that collectively consume 15–20 operator hours per week in many centers. Field results show that AI-driven automation can reduce manual processing time by 60–75 percent, cut transcription and coordination errors by up to 90 percent, and enable sub-minute response times for alerts that previously required 20–30 minutes of manual handling, with payback periods measured in weeks rather than years

Importantly, the presentation emphasizes workflow integration, not replacement: agentic AI operates alongside existing Advanced Traffic Management Systems, SCATS, AVL, CCTV, and traveler information platforms, using APIs and rule-based validation to ensure reliability and auditability. Lessons learned from deployments highlight key practitioner considerations, including phased rollouts with manual fallbacks, clear exception-handling rules, and staff engagement focused on workload reduction rather than workforce displacement. Insights are also grounded in contemporary Road Management Centre operating models, where high volumes of monitoring, reporting, and coordination tasks create strong candidates for automation without interfering with engineering judgment or decision-making

Attendees will leave with actionable guidance on where agentic AI delivers immediate value, how to integrate it safely into 24/7 operations, and how front-line staff can reclaim time for higher-value traffic management and incident response activities while improving accuracy, consistency, and public safety outcomes.