From Miovision, Sajad Shiravi, P.Eng, presented “Traffic Engineering in the Age of Technology and AI” on October 2, 2020.
The current state of traffic management is held back by old infrastructure, undisrupted technology, budget constraints, and legacy systems. Due to these limitations, Shiravi says that incremental improvements and integration with legacy systems is key in the age of technology and AI.
Miovision collects and processes traffic data that helps cities plan and manage mobility using both software and hardware. Their traffic AI uses deep neural networks to constantly learn and improve regardless of changing weather conditions, given enough data to learn. Their Miovision TrafficLink hardware integrates with existing traffic infrastructure and offers “processing on the edge,” where the big video data is processed and only the small analyzed data is sent. The analyzed data involves vehicle and pedestrian detection for vehicle counts and safety such as red light running, pedestrian compliance, and hotspot identification. The data can reveal insights for decision making, especially in unique situations like COVID-19 impacts on traffic in Ontario.
Shiravi says that the traffic industry relies heavily on intersection data collected at spot instances to inform long-term decision making. This involves “signal retiming” where volume data is collected and analyzed for an assumed representative day to generate new timing plans which are set for 3 to 5 years which could become quickly outdated.
Shiravi then explains another approach: Automated Traffic Signal Performance Data (ATSPMs). This approach identifies operational issues and opportunities for improvement by continuously collecting data. The advantages include: measuring rather than modelling; retiming on shorter intervals; and continuous automated measurement of performance of roadways from road users.
Shiravi ends the presentation by saying that the future of traffic and transportation engineering is bright and will be more data driven, technology-centric, automated, and complex. Shiravi believes that as traffic engineering evolves in the age of technology and AI, it will be at the centre of smart city initiatives and open up new careers.
The seminar was presented by the University of Toronto ITE student chapter.
Sajad Shiravi is a Product Manager at Miovision. He is a Professional Engineer in Ontario and has almost 10 years of academic and industry experience in traffic engineering. His main focus is on new sources of traffic data and developing methods and products that turn data into measurable impact. Specifically in the past five years he has been working with traffic signal performance measures helping public agencies incorporate high resolution traffic data into their workflows.