Ever notice a highway change its mind? Well, in Arizona, this occurs every other day. The so-called “flex lanes” on Interstate 17 in the Grand Canyon State are more than a trick of modern engineering and represent a bold example of the impact artificial intelligence and smart traffic flow technologies are having on highways across the United States, with no new concrete in sight.
A cost of $522 million will solve one of its most congested areas in the area around Black Canyon City and Sunset Point in Arizona. This is a two-way traffic road that uses physical separation, overhead gantries, and automated systems to turn traffic around depending on flow. During peak flow when traffic is headed towards Flagstaff, both lanes are dedicated to moving traffic up; however, during days of heavy flow back, it reverses, allowing twice the traffic flow in the southbound direction. The technology used in its systems uses radar and infrared sensors to track traffic flow rate and speed in real time, which are used in algorithms to predict congestion before it occurs.
In Pennsylvania, the data-driven strategy is the same but presented differently. In the area of I-76 Schuylkill Expressway, the physical conditions do not allow road expansion. However, PennDOT has implemented what they call a “Smart Corridor” system. This consists of 72 digital signs and sensors installed along the route that allow temporary use of the shoulders as travel lanes during peak periods. It also features dynamic speed limits that automatically adjust by 5mph increments to coordinate traffic flow and eliminate the “accordion effect” associated with sudden braking. This balances higher traffic volumes at lower speeds while preventing rear-end crashes by braking traffic consistently before the congestion points. In Europe, data analyses on such installations showed 30% decreases in crashes and travel time increases of 5% to 15% faster travel. This is applicable to the Pennsylvania system.
The Georgia “Top End” Express Lanes on I-285 introduce a new element of innovation wherein dynamic pricing is combined with public transit systems. The toll prices fluctuate depending on usage levels in a lane, ensuring that drivers using express lanes can maintain minimum speeds, thereby functioning, in effect, as a “virtual busway” to transport Bus Rapid Transit (BRT) vehicles through designated routes, thus allowing buses to avoid traffic by not requiring separate bus lanes, as in the instance of San Diego’s I-15 managed lanes projects. The Georgia project involves all-electronic tolling, overhead transponders, and dynamic traffic modeling to ensure that express lanes remain free-flowing while normal lanes in that area are reduced to a crawl. The revenue generated from express lanes can be fed back to enhance public transport systems, thus giving a boost to both transport networks.
The technical requirements of reversible and managed lanes are so precise that lane control signals must be visible throughout all lanes with a redundancy in case of misinterpretation. Roadway markings are bi-directional, with median crossovers that include crash-rated barriers that can be opened or closed in a matter of minutes. Artificial intelligence solutions not only incorporate data from traffic sensors but also data related to weather forecasts, incidents on the road, as well as telematic data from cellphones to react to changing conditions. In the state of Texas, AI solutions assess up to hundreds of thousands of miles of lanes in order to detect old signage or infrastructural dangers.
From an economic viewpoint, these smart lanes are high returns on investment infrastructure. By adding capacity without expanding the road, the state does not have the cost of acquiring land or the detrimental effects of construction on the environment. A benefit of the managed lanes in Florida, where the lanes have a barrier-separated design and allow for the ability to directly connect into the lanes, has increased the reliability of freight movement and reduced the delivery times of freight, which has a positive effect upon the economy. A reversible lane system planned in Utah’s I-15 will directly connect into the commuter rail system, making the Wasatch Front a multi-modal backbone. Safety comes first. AI assists with lane control.
This reduces conflict points, takes care of kinetic energy spread over the road, and readies the driver for what is happening next downstream. Other systems such as variable speed limits for Pennsylvania’s roads or congestion-based tolling for Florida’s infrastructure strive for the reduction of sudden stops. These are considered the primary contributors to multi-car accidents. Arizona’s flexible lanes use infrastructure that guards against head collisions while making directional changes. For transport engineers and transport planners, the implication of the deployments is the transition from fixed to adaptive systems. This is because the road system is no longer a fixed piece of infrastructure but an adaptive system that is tuned by algorithms and information. This technology will eventually be integrated with the vehicle-to-infrastructure communication system to make highways capable of sending warnings regarding the area’s speed change, change of lanes, or danger to the vehicle. This foundation work in Arizona, Pennsylvania, and Georgia is solving the congestion problem today and lays the operational foundation for self-driving transport solutions to come.
