ChargePoint Wants To Fix What Americans Hate About Charging Electric Cars

Flow rate, cost and uptime are the most important metrics for electric vehicle (EV) chargers. When on a trip, or at the grocery store, nothing frustrates drivers more than a charging station hampered by either hardware or software issues.

ChargePoint, the biggest EV charging station operator is the United States, is stepping up and taking responsibility for its station uptimes in ways its competition isn't.

The electric vehicle charging provider is investing millions of dollars into its network and information security, including artificial intelligence to bring uptime, the measurement of time an EV charger is fully operational, to near 100 percent, once implemented. ChargePoint says its current uptime is between 96 and 98 percent.

The company defines uptime as the percentage of the amount of individual ports/handles that are active and able to charge on at a given time. The definition is important because some charger companies measure differently.

EVgo looks at "one and done" charging, which is the percentage of time a customer has a successful charging experience on their first attempt. The company says that represents a success metric rather than an availability metric, which lands at 95 percent. Electrify America doesn't share its uptime numbers because it says the qualifications still aren't clear enough to make it a useful comparison.

"At some point, there'll be enough public fast charging EV stations that if a station or two doesn't work, it doesn't have a major effect. But right now, outside of Tesla's network, if a charging station or location is offline, there is often not a nearby alternative. That could mean a costly tow for someone who only had enough juice to get to that station hoping it'd work," Chad Kirchner, VP of Content at EV Pulse told Newsweek.

ChargePoint's Network Operations Center will monitor stations proactively and use predictive analytics and leverage machine learning to find out which chargers are down. The company says this will improve response and repair time. They are also instituting a new comprehensive training program to certify electrical contractors that install EV charging infrastructure.

AI and machine learning will help speed up repairs using social media as an instigator. "Here is a good example. Bob posts on social media something like, 'the ChargePoint station I went to off I-15 in Las Vegas is down.' Sally posts, 'I couldn't charge my car at one of the stations near the Las Vegas strip.' AI will scan and listen for this type of feedback- remembering neither on its own is enough for us to easily ID a problem," JP Canton, ChargePoint spokesperson told Newsweek.

"It will see both, and combine the two points (Vegas near freeway, and Vegas near strip) to pinpoint the area, and check stations – or better yet the exact one where possible- to identify what is wrong. This is particularly useful where there is hardware damage we can't ID remotely, like someone vandalizing a connector."

ChargePoint explains that there is a bit of a grey zone when it comes to software and cellular connectivity that intersect hardware and software. That's where issues can crop up. Overall, hardware is a very small percent of charging failures.

The AI work is in a pilot phase, but ChargePoint is aiming to have it running within six months. It also just rolled out an app to its installers to help in finding and repairing the chargers that need work. Artificial intelligence is meant to anticipate these failures because the company understands that drivers can't always go down the street to another station.

"While most people charge exclusively at home, where to charge when on a longer trip is a big concern for prospective EV buyers. We won't need as many charging stations as gas stations, since charging stations can be placed more strategically. That being said, today is the worst it'll ever be. It keeps getting better," said Kirchner.

"Once other automakers adopt the North American Charging Standard and figure out a universal way to handle payment processing it'll get easier. Right now it's rough if you don't have a Tesla. Hopefully by 2025 it'll get noticeably easier," he added.

The difference here is proactive versus reactive. In the past, a customer could call and report a charger issue, then that call would be acted upon. Now, ChargePoint is aiming to catch problems before an issue is encountered.

"Since all stations are online and connected, the various humans and AI involved can analyze data and predictive analytics to anticipate a failure or detect something immediately when it goes down and flag it for repair, as opposed to waiting on a driver or station owner to reach out, which they don't always do," Canton said.

"Especially for a rural station where the charger isn't frequently used, knowing before someone gets stuck in a real jam is a huge benefit. We can of course repair ASAP, but also pull it off the network so people don't plan on charging there in the first place."