Autonomous ride-hailing fleets need somewhere to recharge between trips, and that’s often a dedicated charging hub rather than a public station. These hubs can look a lot like large dc fast charging sites, but they’re used very differently.
A robotaxi may plug in several times a day, so the equipment accumulates wear far faster than it would at a public site, and every charger that’s out of service leaves vehicles waiting instead of carrying riders. There’s also no driver on board to notice when a session fails or decide when it’s time to head back for a charge.
Those conditions shift the engineering priorities. Components that would last for years at a public site wear out much sooner, and the charging process has to keep running with little driver involvement. The first place that becomes obvious is where the vehicle physically connects to the charger.
Connector life is measured in plug-in cycles
Charging connectors and cables are mechanical components as much as electrical ones. Every connection wears the contact surfaces and flexes the cable, and suppliers typically specify a finite number of mating cycles. At a public station, that count builds slowly over years of occasional use. At a fleet hub, it doesn’t.
A recently completed Voltera EV charging site in Atlanta. High-utilization fleet hubs require charging infrastructure designed around uptime, power availability, and vehicle throughput. (All images courtesy of Voltera)
Mo Dali, director of product and technology at Voltera, explains why that difference changes the way connector wear is managed.
“High utilization shortens connector and cable life because those components are typically rated and warrantied by plug-in cycles, not simply by the amount of energy delivered,” he shares. “A fleet of vehicles charging 24/7 can consume those cycles much faster than a consumer vehicle charging a few times a week. At autonomous vehicle (AV) fleet scale, that makes connector life a measurable operating input rather than an occasional maintenance issue.”
Treating wear as a predictable input means tracking how hard each charger is working and replacing parts before they fail in service. The operating data from these sites also gives hardware vendors a clearer view of what sustained cycling does to their products, which can shape product improvements and warranty terms.
“AV fleets are establishing a new reliability standard for charging equipment by demonstrating what sustained, high-throughput use demands in practice,” he adds.
Keeping failure from spreading
Repeated plug-in cycles accelerate wear on charging connectors and cables, making service life an important design consideration for fleet hubs.
With parts wearing out on a shorter timeline, hub design has to assume individual components will fail. The cables and connectors that wear fastest are among the most common failure points, along with power modules and contactors.
Planning for those failures involves both how the site is designed and how quickly a failed part can be replaced.
“Hubs are designed so a single equipment failure does not become a fleet-wide problem, with a clear spare parts and technician strategy supporting that design,” Dali explains.
Depending on the size of the hub and how it’s used, critical replacement parts may be kept on site or elsewhere in the same market, with trained technicians positioned nearby. Real-time monitoring that extends past the chargers into the site’s high- and low-voltage systems helps crews locate a fault quickly while the rest of the hub keeps operating.
Megawatt loads and storage
Dozens of high-power dc chargers running around the clock add up to a load that’s closer to an industrial facility than a typical commercial site, which makes available utility capacity a major factor in planning a hub.
“A large AV hub typically requires approximately 7 to 10 MW of power, although the exact load depends on fleet size, charger selection, vehicle dwell time and operating profile,” Dali says.
Onsite battery storage can help when the grid can’t deliver enough capacity or when peak demand charges make the investment worthwhile. It isn’t a simple fix, though. Many AV hubs are in dense urban areas where batteries and solar add capital cost and take up space that could otherwise go to chargers and vehicle circulation.
“For that reason, storage is not automatically the best solution,” he says. “It has to be evaluated site by site against the available utility capacity, operating needs, space and economics.”
Large fleet charging hubs, such as this site in Houston, can place multi-megawatt demands on local electrical infrastructure while also requiring space for vehicle circulation and charging equipment. That makes available utility capacity a key consideration early in site planning.
Coordinating the session
A person still generally connects the vehicle today, even as the industry develops automated approaches for plugging in and other routine tasks. The rest of the session, however, has to work without a driver confirming that the right vehicle is on the right charger or that charging actually started. Those checks still have to happen, even with no one onboard to make them.
“Removing the driver does not remove the need for precise coordination,” he says. “The vehicle, charger and supporting software systems must communicate in lockstep to confirm identity and availability, initiate the correct charging sequence, monitor the session and share status in real time.”
Deciding when a vehicle should return to the hub is part of that coordination. Fleet operators weigh each vehicle’s state of charge and location against expected rider demand. Live data on charger availability and site capacity lets routing respond to what’s actually happening at the hub instead of relying on a manual check or static schedule, which helps cut deadhead miles and keeps vehicles from arriving at a queue.
Diagnosing sessions
When a session fails at a public station, the driver usually notices right away. At an AV hub, attendants are on site to connect vehicles and support operations, but a failed session still has to be caught through continuous monitoring. If a vehicle is plugged in but charging doesn’t proceed as expected, the facilities team can remotely review the communication handshake and fault codes to see where the session broke down.
From there, it comes down to figuring out which side of the connection the problem is on.
“By comparing how a vehicle performs across different chargers, or how different vehicles perform on the same charger, we can often determine whether the issue sits with the vehicle, the charger or the communications layer,” he explains.
Fully automated charging is the next step for these hubs, and Dali notes that it will depend on the entire hardware and software ecosystem working reliably together.
“Automation only creates value if the full process works consistently without adding dwell time or operational complexity,” he says.
Filed Under: Charging, FAQs, Featured Contributions