Simanaitis Says

On cars, old, new and future; science & technology; vintage airplanes, computer flight simulation of them; Sherlockiana; our English language; travel; and other stuff


IF I AM to believe predictions, we will all be driving, er… make that non-driving, autonomous battery electric vehicles in a few years. Yesterday here at SimanaitisSays, counter arguments were offered from Automotve News and the Institute of Electric and Electronic Engineers that autonomous BEVs make no sense. Today, let’s look at evidence from Wired magazine, Automotive News, and Hyundai.

Autonomous BEVs? Battery electric vehicles are already range-limited with this liability compounded by their operation in hot or cold climate. What’s the implication of adding autonomous-vehicle power drain?

Wired magazine, February 6, 2018, states that “Self-Driving Cars Use Crazy Amounts of Power, and It’s Becoming a Problem,” by Jack Stewart.

The first autonomous prototypes back in the late 1980s were vans, their volume needed to fit computers of the day. Today’s autonomous-vehicle hardware has mitigated this, but the subsystem is still an energy hog.

Wired notes, “All the data needs to be combined, sorted, and turned into a robot-friendly picture of the world, with instructions on how to move through it. That takes a huge computing power, which means huge electricity demands. Prototypes use around 2500 watts.”

Wired quotes Wilko Stark, Mercedes-Benz vice president of strategy: “To put such a system into a combustion-engined car doesn’t make any sense, because the fuel consumption will go up tremendously.” On the other hand, Wired notes, “Switch over to electric cars, and that draw translates to reduced range, because power from the battery goes to the computers instead of the motors.”

“Maybe,” Wired says, “you’re old enough to have dealt with a parent who turned off the car’s AC to save gas. Now imagine having to turn off the self-driving abilities just to make it to your destination without running out of electrons.”

Chip Makers on the Case. Wired notes that Nvidia is working on a computer platform consisting of “two Xavier [new highly efficient] chips and two more GPUs” that can perform “320 trillion operations per second and keep power consumption to an acceptable 500 watts.”

Fuel Cells to the Rescue? Automotive News, February 26, 2018, reports “Hyundai: Fuel Cells Best Bet for Self-Driving Cars.”

In his analysis, Asia Editor Hans Greimel shares the Hyundai view that BEVs, with their relatively small battery packs, can not supply the necessary juice for full autonomous operation. But fuel cells generate their own electricity and do so extremely efficiently.

According to Hyundai engineers, full autonomy can consume as much power as 50 to 100 laptops. As for the fuel cell’s advantage over a BEV, the Hyundai Nexo generates more than triple the output of the company’s Ioniq battery electric.

This Hyundai Nexo fuel-cell crossover, driven around the Olympic Village at South Korea’s Winter Games earlier this year, was equipped with a Level 4 system, one level short of all-condition autonomy. A non-autonomous fuel-cell-powered Nexo comes to the U.S. in the fourth quarter 2018.

Greimel quotes Kim Sae-hoon, vice president of Hyundai Motor Group’s fuel cell group: “If we get a perfect autonomous world, then the vehicle will need a lot of energy for computing. We think hydrogen can provide a beneficial platform.”

There’s still that “if.” But I believe autonomous BEVs won’t be part of it. ds

© Dennis Simanaitis,, 2018


  1. jlalbrecht64
    July 6, 2018

    I’m not a Luddite. I actually love new technology. In my semi-expert opinion, the killer for autonomous vehicles is not the power drain. Computers will continue to draw less and less power and batteries will continue to improve (and hopefully someone will make a great discovery and a fortune finding a denser energy solution). The problem is the logic behind autonomous vehicles.

    I posted two comments about this subject almost exactly a month ago here:

    View at

    and here:

    View at

    A very short summary and a couple of additional comments:
    – Cars have had windshield wipers 100 years. We’ve been producing “autonomous” wipers for 40+ years. And yet we still don’t have wipers that can handle a light drizzle well.
    – The issue of liability is THE key point of autonomous vehicles. Autonomous cars will kill people. Who will be responsible? Right now it is clear who is responsible. If it ends up in the future that, “no person is responsible because it is the product of a faceless corporation” then a) I’ll stay off the streets where these things are driving, and b) there are going to be some serious legal fights which may kill the technology through bad press or massive financial judgement. Autonomous cars are not going have a big enough potential market soon enough to survive major litigation (IMO).

    I didn’t put this anecdote in my comments, but I write also from personal experience. Every car I’ve owned for the last 18 years or so has had autonomous wipers. My current care is an €80k 2013 BMW that has the (at the time of purchase) top-of-the-line sensor-driven wipers, and they still aren’t perfect.

    • simanaitissays
      July 6, 2018

      A very interesting point. As you know, I’ve earlier raised the liability/ethical issues of vehicle autonomy here. To me, the power aspect is just one more shortcoming.

      • jlalbrecht64
        July 6, 2018

        “A very interesting point. As you know, I’ve earlier raised the liability/ethical issues of vehicle autonomy here” – I didn’t remember that! Since I read most everything you write, maybe I read what you wrote and internalized it before I made my comments.

        I own a software and consulting firm. We work on large industrial sites, but our products are only software. Despite that, our clients require that we keep high liability insurance. The highest requirement we have is something in excess of € 6.5 million (IIRC) and a huge chunk of that is for personal liability. That insurance is very expensive (many thousands of €/year) and dependent on the number of employees. Very expensive and are small and have never had an incident with our software in nearly 24 years of business.

        So I’m wondering how that autonomous automobile software liability will work out once robot cars start running people over. Or when autonomous vehicles start crashing into expensive assets. What are insurance companies going to do when they realize that a poorly written algorithm from car company X installed in 100,000 vehicles puts them on the hook for potentially billions of dollars of settlement claims?

  2. sabresoftware
    July 10, 2018

    Electric vehicles and autonomous vehicles. Two subjects that this “Luddite” is increasingly amused by.

    Start with electric vehicles. The first electric vehicle was developed as early as 1832, with practical versions around 1870. The reason that gasoline powered cars took over is primarily energy density (25.8 times by weight, 7.92 times by volume for gasoline over electricity based on current highest density battery technology). To get similar range in an electric vehicle requires large (heavy) batteries, which exact a penalty in increased energy consumption.

    Range is another major factor. I can drive From Edmonton to Vancouver in about 12 hours, with a couple of fuel/rest stops and driver changes. A similar trip in an electric vehicle could take several days with overnight charging stops. And if I get my math wrong and run out of gas, an auto club refuel is available. If I make the same mistake in the electric car I guess I’d need a tow.

    The biggest flaw with trying to convert our transportation fleet from hydrocarbon fuels to electricity is that the electric grid and generation system would need to be increased by almost 70%, but probably a little less, as we’d still need fossil fuels for jet transportation. As 63% of US electricity is generated from hydrocarbons (23% in Canada), increased electrical demand to replace hydrocarbon consumption in vehicles would probably just increase hydrocarbon consumption for electrical generation.

    Most renewable generation alternatives are restricted by geography. Hydroelectric requires the necessary water sources and also meets a lot of opposition these days when new projects are proposed, especially large scale ones. Wind turbines depend on having prevailing, steady, wind conditions, which limits locations. Wind turbines also have opposition due to bird kills and complaints from nearby residents about pressure wave effects. Solar needs to have good access to the sun, so many locales are not optimum, and large scale solar farms take up significant real estate.Wave energy and geothermal also need specific geographic locations. Because of the geographic location requirements, significant transmission infrastructure is required to get the generated electricity to where the people are. Transmission losses over long distances can be significant, and transmission lines are vulnerable to wind and ice storms. And a visual eyesore. Nuclear has its issues, and in North America no new plants have been built about 25 years.

    I’m not saying that we continue to burn as much petroleum and coal as possible, but just saying that a wholesale shift to electricity is not realistic, particularly in the short term.

    Autonomous vehicles pose a number of concerns, including of course liability issues, and reliability issues. On two vehicles that I have owned that have rudimentary driver aids (sensors and cameras), I have had proximity sensors that would quit working if too much ice built up over the sensor, and a rear view camera that periodically decides not to work. In both instances as the computer brain receiving the sensor/camera information was mine, I am able to compensate for the data loss. What happens to an autonomous car when this happens? Does it just stop dead in its tracks?

    Lane keeping technology. What happens when road markings have faded or are obscured by snow/slush/mud?

    What happens when I want to drive down a road that isn’t on google maps yet (frequent around here with the ever increasing residential developments)?

    What about GPS accuracy that has me several yards away from where I actually am? Does the system drive by where it thinks I am, or by the actual roadway as defined by local sensors? What happens when the sensors cannot properly define the roadway due to lack of lane markings, or even properly defined edges (i.e. gravel road or road under construction)?

    How does the system navigate a parking lot?

    What about dealing with all the non-autonomous entities out there (actually most of them ARE the autonomous ones), such as kids or pedestrians in general, non-autonomous vehicles driven by total morons, non vehicular risks like a tree falling into a roadway, wildlife, etc., sudden appearance of a road collapse, and if nothing else pot holes?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

<span>%d</span> bloggers like this: