The milestones for self-driving cars, first in the form of robotaxis, keep being recorded. As the mainstream arrival of the technology gets closer, and the vehicles are almost entirely electric, the question of how they will charge and how they will plug in grows in relevance.
Fleets of cars which return to special charging depots can be plugged by a human “plug jockey,” and in fact if the lot is large this can be reasonably cost effective, as the human can be kept working and the cost per plug/unplug reasonable. That is how it’s done now and how it will be done in the early years of such charging lots. In time, though, there are a lot of reasons to want automatic plug-in and charging.
Automatic plug-in allows cars to travel to charging when they need it, or it makes sense. That can be true not just for robotaxis, but also even private cars, even if those private cars are not capable of driving a person around at full speed on daytime city streets. A self-charging car can make short trips to nearby charging lots on quiet streets (for example at night) and at low speed, making it easier to do this safely. This allows the ultimate electric car, one that is simply always full — and you’re not even sure how and when it does it. While today, many car buyers worry about how they will charge an EV, a car that charges itself creates a magic experience not even possible with gasoline, and without any special fast charging.
Self-charging can also make sense in parking lots, particularly in offices, so you don’t have to build lots of charging stations. The cars can share them. In general, self-charging cars mean we don’t have to build a massive charging infrastructure at all. Instead, automatic stations, located near power substations can serve most people, including those who don’t own their own parking space and can’t install charging.
Cars that can travel to charging also have the ability to do opportunistic charging when energy is cheap, as it can be from 8am to 2pm in an area with a lot of solar, or any other period of surplus power. This is not just a financial win for the car owner, it’s a way to absorb surplus power and deal with the variability of renewable sources on the grid. Cars don’t much care when they charge (as long as they are not in use) and so they are the natural partner of a grid with variable supply. For human driven cars, they need to make sure they park plugged in any time that surplus may be available, but self-charging cars can just go if they are free, and disconnect and return if they are needed.
MORE FOR YOU
How will they plug in?
In these early days, they will need a human. But there are several ways that plug-in might be automated, which allows you to have smaller, low-volume stations and lower costs.
Normally, the core rule is that there are many more cars than there are non-home chargers. That means you would want to put the expensive stuff in the charger and not in the car. That rule would even imply we should not have bothered to put internal chargers in all the cars as we do today. (The internal charger runs on household AC electricity at 120 and 240v, and it’s what does the “level 2” slow charging. While people often call the box they have in the home, or in the trunk of the car, or in the parking lots a “charger” it actually isn’t — the charger is built into the car, and the other device is really just an expensive plug and extension cord, technically called an EVSE. A fast charger uses DC and is really a charger. We needed internal chargers in the cars so everybody could charge easily at home and so that the early public EVSEs would be cheap to build. Today, the internal charger is not very expensive, and so it’s going to be there for good.
The above rule would suggest automatic plug-in should be done by a robot plug jockey which can hold and plug in a standard cord. That would be true if it weren’t for the fact that a self-charging car is already a fairly capable robot and able to do most of the moving. If charge ports were located on the front and back of cars, then robocars could plug into a fixed plug mounted at the right height, making the station install dirt cheap.
Reality interferes in a few ways. First, the port is almost never on the front or back, and if it were, it would be damaged in accidents (especially on the front) increasing their cost. Secondly, cars tend to mount ports at different heights, and they would need to standardize a height. A plug with just one actuator to adjust the height would be fairly low cost, but again, plugs are not on the front or rear. For existing styles of ports, the plug would need to move up and down and also in and out — the car could do the rest of the moving. Ideally all plugs would try to be at a standard angle.
Charging Robots
That’s still much less fancy than the special robots that have been built by companies like Dutch company Rocsys which recently received a funding round. Their large and more complex robot is meant more for regular cars in fleet yards or in places where getting out of a vehicle to plug in is problematic. Tesla also famously showed a “snake” robot for plugging in their cars many years ago, and several other companies have demonstrated these plugs. At present they tend to be quite expensive and not designed to take advantage of a car that can position itself precisely. Hyundai has also shown off a corded charging robot prototype known as the ACR.
In addition to charging robots for fixed stalls, there are charging robots that move and have batteries. These can move around a parking lot to charge cars, which is very useful for cars unable to move themselves. The NaaS charging robot is such an example. Unfortunately, there are losses involved in charging and discharging from a different battery, and it also puts wear on that battery which has a cost, so this is unlikely to be your everyday solution.
A new connector
There was just a big battle among connector choices between Tesla’s NACS and other systems like J1772, CCS and CHAdeMO. NACS appears to have won. So it would seem odd to propose a new connector. A new connector would not really be in competition with the NACS because a fleet car might have its own system and never use public chargers. If the car is going to roam it would need at least an adapter.
An adapter would not be an option for a connector on the bottom of a vehicle, though there is merit to that location. While anything there would need to be protected from weather and road debris, it’s nice because it allows charging stations to just be a box on the ground that a car drives over. With suitable spring action, the connector could work as a physical rather than actuated device, to keep costs low.
Wireless charging
For years, many companies have been promoting inductive charging for consumer cars. For this, the car has an inductive coil mounted in the bottom and another pad with such coils is placed on the ground. The car simply drives on top of the pad. This is mostly marketed to human drivers as convenience — you don’t need fully precise positioning, and you just drive onto the pad and walk away, without even the time of plugging and unplugging.
This approach has not yet gotten too much traction for a few reasons. The charging pad and coils can be a significant cost add-on for existing vehicles, and no vehicle has put them in standard. There is no standard yet for this, so vehicles would need to also have a regular connector. Inductive charging has some losses (though AC charging with the internal charger also has losses.) Inductive charging uses AC that must be converted to DC in the car — for high power charging that can involve some fancier gear in the car.
While vendors claim the losses of inductive charging are similar to conductive plugs, even a small difference adds up over the life of a charging station. However, for robocars, it’s possible to put the elements nearly in contact due to precision parking, based on the lowest clearance car being handled by a station.
Tesla recently acquired a wireless charging company named Wiferon, so it may have ambitions in this area.
Swap
While battery swap has many issues, for cars that have it, it does have the virtue of not needing a human, so it can also be used for self-charging cars. The car must be specially built for swap, of course, with a different battery pack design. Tesla’s structural pack design would unlikely be compatible, for example.
Vehicle To Grid
V2G, as it is known, is not yet a mature technology. With V2G, it is hoped that cars can provide power to the grid when it needs it, and the price paid for this is high, enough to compensate for the wear on the battery and the reduced range of the car after a session. But one problem with V2G is the cars have to be plugged into a two-way charging station able to handle V2G during those special moments (mostly hot summer evenings) when the grid is running short. Here, the robocars have a major win — if the summons come that the grid is paying very well for power, they can drive to a two-way location to sell their power. In this way, you don’t need to build that many two-way stations in the hope of catching a car in one during a grid shortfall. Currently two-way stations are expensive, though in the future, homes with solar and inverters may be able to do this more cheaply.
Indeed, since shortfalls that need V2G come only at certain times, human plug jockeys are a reasonable solution here, if the cars can make the trip.
