In December 2023, RM3P Digest chatted about the Artificial Intelligence-Based Decision Support System (AI-DSS) with Kamal Suliman, then VDOT’s NoVA District Traffic Operations Director and AI-DSS Steering Committee Member, and with Mike Clements, VDOT Assistant Division Administrator for Corridor Management and AI-DSS Future Steward. Excerpts from these conversations are presented below.
What does AI-DSS mean for staff at VDOT’s Transportation Operations Centers (TOCs)?
Kamal Suliman (KS): I see AI-DSS as both input and output – input based on what the system is designed to perform and output being the Operators’ perspectives on decision-making on the data and information gathered or provided to them. In today’s environment, when there’s an event [identified by the] VSP CAD dispatching system, we would have a list of response plans [and a] list of cameras and DMS within a prescribed area; [we would also look] at Google, Waze, and the information dispatched from different places. [The AI-DSS] would provide [important] recommendations [to] operators.
Based on the current AI-DSS milestones schedule, which ones would you deem critical to watch out for?
KS: My preference is always to go with simple things that can be verified manually, but if you deal with the most difficult part, it’s going to be so many variables that we [may not be able to] figure out if the system is operating as expected or not.
Recently, the RM3P team has been hard at work on defining long-term “Sustainability” among partnering agencies. What does Sustainability mean for AI-DSS and the Commonwealth’s corridor program? Also, as a future steward, please share your vision on what AI-DSS in the future might look like?
Mike Clements (MC): Generally, Sustainability is maintaining effectiveness or value of something into the future. Sustainability in regard to RM3P is about taking some actions that are necessary to ensure that the AI-DSS and [and the other RM3P program elements] will be able to function effectively into the future. There are two actions that we might take: determining who has primary responsibility or stewardship of these RM3P initiatives after initial deployment, which we know already; and then the second piece of it is identifying long-term funding streams for operations and future enhancements and technologies. The stewardship responsibility for the AI-DSS, after deployment, is going to transition to the Corridor Management Group that I manage under the Traffic Operations Division. So as a future steward, the Corridor Management Group will continue to maintain those stakeholder relationships that have already been built, look out for future innovations and upgrades to the system, and seek funding for the next generation of the Decision Support System. We are also going to need to periodically review and revise the strategic plans to help us stay focused and able to adapt to the new technologies or trends that happen over time.
What AI-DSS function or functions will be especially important for regional operating agencies? Walk us through some of the innovative technologies scoped for AI-DSS that catch your interest.
MC: I think one of the important functions is going to be the multi-modal response plans that are going to help us do more than just detour traffic. Availability of the plans will provide us the opportunity to change modes between transit, carpooling, walking, biking, telecommuting, as well as [selecting] a different route. As far as some of the innovative technologies are concerned, obviously it’s the prediction of congestion, incidents, and parking…[that] will be [especially] interesting. This tool should help regional agencies to better respond/prepare to mitigate events and it may even stop some events from occurring. This tool will allow us to be more proactive in managing incidents and congestion, as we would get a chance to look at things ahead of time. [The] last thing is having faster, better, more targeted communications among partners and agencies, since we are operating from a shared set of data and information.
What are the key benefits to regional operating agencies that the AI-DSS is expected to provide?
KS: Assuming that everybody is going to be able to have access to [AI-DSS] information and actively participate in the value that it provides, …there will be a common understanding that stitches all our players into a single platform with all [the] different modes. For example, Metro could shut down and [Operators] might have no idea [until the situation] starts to pose a problem. So having that kind of knowledge would be the biggest benefit.
MC: I think the biggest benefit is going to be the large, holistic view of the entire system now that we [will] have a dataset that includes roadway, transit, and parking data. Just having that holistic view [and being] able to see it all [in the form of] a one-stop-shop for the data [will] probably be one of the bigger benefits.
What roles and responsibilities do you think regional stakeholders should have in the deployment and operation of the AI-DSS?
MC: I think [stakeholders have] got to support and facilitate the interaction between the AI-DSS team and the agency operators and subject-matter experts, and try to make sure that the information is going both ways; [they also must confirm that] the needs and interests of those agencies/stakeholders are being met. So there’s a supporting/facilitating representation kind of role for stakeholders. I think they are also advocates for participation to make sure that the regional agencies remain informed about RM3P/DSS and its capabilities.
KS: I think that the most important role is the validation [by the agency] of what is going in and what is coming out. If [the agency is] engaging in the process of validation or checking-in on a routine basis for the initial deployment to make sure it’s still doing what it’s supposed to do, we will win over [regional stakeholders’] confidence in the system. Engagement is pretty important and the key to that is having [timely and] good quality [information].
What is your vision about how RM3P, and AI-DSS in particular, will function and perform over time? What improvements do you anticipate in performance of the regional transportation network?
KS: Hopefully learning more and getting better will be part of the exercise. It is important to set the right expectations, make incremental improvements, and deliver useful and good quality solutions so as not to lose interest among stakeholders. I recommend [that] the team make [continuous] improvements around different scenarios and incidents.
MC: Getting real-time or near-real-time data out of the systems on multiple infrastructures [will] allow for much more coordinated response between local and regional agencies. It [will] give us better ways and multiple options [for] coordinating with each other to solve problems.