Toll Insight spoke with Sean Qian, Director, Mobility Data Analytics Center (MAC).
1. Please tell us about yourself and the Carnegie Mellon University Mobility Data Analytics Center.
I am a Professor of Civil and Environmental Engineering at CMU. My research interests lie in the integration and optimization of civil infrastructure systems. The primary focus is to manage an aging and overcrowded infrastructure and cyber-physical-social systems and to build a sustainable and resilient infrastructure network. I have been developing optimization and statistical models of infrastructure system management by applying theories of AI/ML, data analytics, network flow and economics.
The Mobility Data Analytics Center (MAC) is a research facility at CMU that aims to collect, integrate, and learn from the massive amounts of mobility data and develop smarter multi-modal multi-jurisdictional transportation systems. The ultimate objective of MAC is to:
Provide archived and real-time traffic data of every element of multi-modal transportation systems.
Reveal the behavior information for both passenger transportation and freight transportation.
Serve as a key instrument for managing transportation systems.
Target a range of users including legislators, transportation planners, engineers, researchers, travelers, consultants, and technology companies.
Over the years, MAC has been developing research studies sponsored by USDOT/FHWA, US DOE, NSF, state/local agencies, Honda, Fujitsu, IBM, and non-profit organizations.
Most recently, I founded a CMU technology spinoff firm, TraffiQure Technologies, to commercialize technologies developed through MAC research projects.
2. What are your research focus areas, especially as related to tolling?
I have been leading the efforts developing mesoscopic network flow models in large-scale transportation networks. It models second-by-second traces of passengers and vehicles in general networks of roadway, parking, public transit and mobility services. The key is to leverage multi-source data to understand travelers’ choices of routes, parking, destination, time, and modes, which may be influenced by travel time, cost (e.g., tolling), safety and reliability. Our team has developed high-granular network models for assessing the system-level impact of congestion pricing, bridge tolling, mileage-based charges, TNC pricing, curbside pricing, and various pricing strategies, as economic instruments to manage transportation systems. Those system-level impacts include traffic delay, VMT, VHT, reliability, accessibility to essential resources, emissions, energy consumption and safety.
3. Please summarize your recent TNC incentivization study.
As usage of ride-hailing services continues to grow, transportation network companies (TNCs), such as Uber and Lyft, start to have increasing influence on network traffic patterns. As cities attempt to reduce the number of cars on the road and ultimately emissions at peak travel times, they also struggle to develop inexpensive and equitable solutions to improve transportation network performance. We propose an Optimal Ride-Hailing Pricing (ORHP) system that could address these concerns.
Through ORHP, a subsidy is provided to TNCs in exchange for guaranteed improvement of the network as it pertains to route selection by TNC riders. Public agencies can set a surcharge or credit for selected roadway segments that would be used by non-private driving TNC vehicles. A surcharge or credit will be issued to TNCs each time a TNC trip uses any of those roads. With those (dis)incentives, TNCs would determine how to route riders along with some compensations that would benefit TNC’s fleet as a whole. Riders are provided by TNCs with multiple route options, and incentivized to take a ride that may deviate from the shortest possible route through lower fares or other compensations offered by the TNC. This system is based on the premise that a TNC fleet, once reaching a critical market penetration, oftentimes a small fraction, could have influence on general traffic flow patterns and serve its own best interest in improving fleet efficiency. TNC riders can voluntarily participate in choosing deviated routes with compensation, which is less controversial than other pricing strategies, such as congestion pricing. While the decision ultimately falls upon the rider to take a longer route that will assist in improving traffic congestion, both TNC and public agencies stand to benefit from a service and profit standpoint. Public agencies can save from capital investment of pricing, as fare and sensing systems are already an integrated part of the TNC mobility services. Total costs could be reduced for TNCs as fleet vehicle travel time can be better optimized through this system, and the financial benefits of the subsidies could be shared by both riders and service providers. Even a relatively small investment was shown to yield promising outcomes.
For more information, go to:
4. How do you bridge academia to industry via your commercialization efforts?
My passion is to deploy start-of-the-art technologies, to help solve real-world problems. From the research perspective, I love spending time discussing with various stakeholders and deployment partners to understand their real-world needs. By overlaying those issues and problems with my expertise, I look for technology-based solutions and research the feasibility of those technologies through pilot studies. If those studies are promising, then those solutions can be moved to a deployment stage where I actively communicate with public agencies and industrial partners to seek deployment opportunities. I firmly believe that academic, industry and governmental agencies should work together to identify the best solutions for our society. Our transportation industry could use more innovations, and our academic community could benefit from transferring technologies addressing real-world problems for a broader impact.
TraffiQure Technologies was formed in 2020, spinning off technologies resulting from a decade of transportation and Artificial Intelligence research at the Mobility Data Analytics Center. TraffiQure Technologies is integrating machine learning (ML), artificial intelligence (AI), and large-scale system simulation technologies into transportation engineering/planning leveraging multi-source multi-jurisdictional data. It has exclusive licenses from CMU to commercialize:
AI-powered proactive incident management (24/7 incident prediction and proactive actions)
Sensorless parking management, pricing, and enforcement
Shared mobility service platform for under-served communities
Social-digital-twin for network-wide decision-making
5. As a university professor, which careers interest your students and how do we bring them into tolling?
My students come from various backgrounds, e.g., Engineering, Economics, Public Policy, Computer Science, etc. Many of them are interested in developing Intelligent Transportation Systems including tolling infrastructure, strategies, and management systems. I think the tolling business is particularly interdisciplinary, which can be open to students from various backgrounds. The key to attract students into tolling is to showcase the societal impact of tolling in potentially improving mobility, safety, economic competitiveness, and equity of our transportation systems. We need to help students identify which specific roles within the tolling industry align with your student's skills, interests, and career goals. Some common career paths in the tolling business include transportation planners and operators, toll system technicians, economists, software developers, customer service representatives, project managers, data analysts, and policy analysts. The tolling industry is constantly evolving, with the integration of advanced technologies and the pursuit of more sustainable practices. We would need to encourage students to stay updated on industry news and emerging trends.