Toyota Looks To Artificial Intelligence For Next-Generation Battery Technology

3 weeks ago by Steven Loveday 27

Toyota Prius Prime

Toyota Prius Prime

It turns out that Toyota will begin employing artificial intelligence (AI) related to electric cars. However, not in the way most might think.

Now that Toyota is moving forward in its development of longer-range EVs, instead of fuel-cell vehicles (FCV), the company is eager to be the first to discover some “hidden” battery range secrets.

Toyota Mirai

Toyota Mirai

Toyota will invest $35 million into its Toyota Research Institute (TRI), over the course of the next four years, specifically to use AI to search for the answers to increasing battery range. CEO of TRI, Eric Krotkov, shared:

“We want to accelerate the rate at which we can design or discover new materials for fuel cells and batteries.”

The AI will be put to the task of searching for new materials that can better power EVs and FCVs. Toyota believes that using technology to find the answers could put them years ahead of the competition. The company is also set to begin a multi-faceted partnership with a U.K.-based “accelerated materials innovation” firm (Ilika Plc), along with Stanford University, and the Massachusetts Institute of Technology. Toyota is also willing to expand this partnership to multiple other contributors.

TRI is also working toward autonomous vehicles. However, the company was specific about its intention to market toward the elderly, creating vehicles that will allow older people to “age in place,” instead of having to rely on others, or be confined to an elderly care facility. Retired U.S. military robotics expert, Gill Pratt, was brought on board Toyota’s team to head up this venture.

Toyota hopes that with this major step, it can reach its goal of having a carbon footprint that decreases by 90 percent by 2050. Though Toyota is still pursuing FCVs as a substantial part of this role, company president, Akio Toyoda, admitted that the automaker is now shifting immediate focus to battery electric vehicles.

You can read Toyota’s official press release below:

$35 Million to Accelerate Materials Science Discovery  

Projects will apply artificial intelligence and machine learning to speed development of materials for next-generation energy

March 30, 2017

Palo Alto, Calif., March 30, 2017 — The Toyota Research Institute (TRI) will collaborate with research entities, universities and companies on materials science research, investing approximately $35 million over the next four years in research that uses artificial intelligence to help accelerate the design and discovery of advanced materials.  Initially, the program will aim to help revolutionize materials science and identify new advanced battery materials and fuel cell catalysts that can power future zero-emissions and carbon-neutral vehicles.

“Toyota recognizes that artificial intelligence is a vital basic technology that can be leveraged across a range of industries, and we are proud to use it to expand the boundaries of materials science,” said TRI Chief Science Officer Eric Krotkov. “Accelerating the pace of materials discovery will help lay the groundwork for the future of clean energy and bring us even closer to achieving Toyota’s vision of reducing global average new-vehicle CO2 emissions by 90 percent by 2050.”

Initial research projects include collaborations with Stanford University, the Massachusetts Institute of Technology, the University of Michigan, the University at Buffalo, the University of Connecticut, , and the U.K.-based materials science company Ilika. TRI is also in ongoing discussions with additional research partners.

“This represents a fantastic opportunity to drastically advance the use of databases and machine learning methods in materials discovery,” said Jens Norskov, Professor at Stanford University and director of the SUNCAT center.  “The partnership combines theory, computation and experiment in an unprecedented, concerted effort. We are particularly excited by prospects for an avant-garde approach to catalyst development for fuel cells.”

Research will merge advanced computational materials modeling, new sources of experimental data, machine learning and artificial intelligence in an effort to reduce the time scale for new materials development from a period that has historically been measured in decades. Research programs will follow parallel paths, working to identify new materials for use in future energy systems as well as to develop tools and processes that can accelerate the design and development of new materials more broadly.

In support of these goals, TRI will partner on projects focused on areas including:

  • The development of new models and materials for batteries and fuel cells;
  • Broader programs to pursue novel uses of machine learning, artificial intelligence and materials informatics approaches for the design and development new materials; and,
  • New automated materials discovery systems that integrate simulation, machine learning, artificial intelligence and/or robotics.

Accelerating materials science discovery represents one of four core focus areas for TRI, which was launched in 2015 with mandates to also enhance auto safety with automated technologies, increase access to mobility for those who otherwise cannot drive and help translate outdoor mobility technology into products for indoor mobility.

 Source: Toyota

Tags: , , , , , , , , , , , ,

27 responses to "Toyota Looks To Artificial Intelligence For Next-Generation Battery Technology"

  1. Hugh says:

    Computer technology like IBM’s Watson can be used to gather and compile large amounts of data but I don’t personally see how AI can be used in materials sciences. But Hey, Whatever Works!

    1. Pushmi-Pullyu says:

      It’s ridiculous how often the term “Artificial Intelligence”, or A.I., is misapplied to expert systems software. Anyone who actually understands how very far computers and computer programs are from true intelligence, actual sapience or consciousness or cognition, must laugh to themselves and/or roll their eyes every time someone tries to paste the label “A.I.” onto such a software program or suite of programs.

      1. Timmy says:

        It’s gotten much worse just in the past year, when “AI” has (awkwardly) entered the vernacular.

      2. ItsNotAboutTheMoney says:

        Expert systems are a branch of AI.

        The problem is that people are (somewhat ironically) narrow in what they think of as AI.

        1. Pushmi-Pullyu says:

          No, the problem is people using a ridiculous amount of hype in trying to paste the label of “intelligence” on computer programs which are not intelligent to even the smallest degree. I am aware of what the Etymological Fallacy is. This ain’t it.

          Robotic intelligence has reached the approximate level of a middling-smart bug, and is struggling to reach the level of smarter bugs. Expert systems don’t try to use even that very limited level of actual intelligence; they’re just better than a more general-purpose program at identifying and allowing choices among a range of pre-programmed options within a very narrowly focused subject.

          Calling an expert system program “intelligent” or “Artificial Intelligence” is just as silly as claiming that your word processing program can actually understand what you’re writing because its SpellCheck can identify spelling errors and possibly a few grammar errors.

          A truly intelligent computerized or robotic system would have some degree of understanding of what the data it’s processing actually means. Expert systems do not, not even to the smallest degree; no more than less sophisticated software does. Expert systems are no more resistant to the GIGO problem (“garbage in, garbage out”) than less sophisticated programs, for the same reason: They have no understanding at all of what the data actually means.

    2. ItsNotAboutTheMoney says:

      They’d be using computer modeling and AI techniques to have software try to find better ways of doing things.
      The idea of the use of AI is to help avoid having to try all permutations.

      This kind of stuff might be relevant:
      https://www.ted.com/talks/maurice_conti_the_incredible_inventions_of_intuitive_ai

  2. Get Real says:

    Methinks that Coyota should look to Artificial Intelligence to realize that the PHYSICS of Hydrogen will never change and to give up that boondoggle.

    Maybe then they will start a serious PEV program if its not to late.

  3. electron says:

    with hydrogen being the most abundant element in the universe, and on earth, rising sea level could be used as fuel..we just need the innovation to split the H2O in an energy efficient way….hydrogen fuel cells are a good long game solution, but in the shorter term batteries will be cheaper…..use your solar panels to make liquid hydrogen fuel, and resupply your oxygen tanks for your home oxygen bar

    1. Pushmi-Pullyu says:

      “…we just need the innovation to split the H2O in an energy efficient way…”

      Please explain what sort of magic you plan to use to change the physical properties of the element hydrogen, so that it’s not hopelessly hard to deal with, so that it could possibly someday become a practical everyday fuel that doesn’t involve massive losses of energy and ridiculously expensive infrastructure.

      Certainly there’s no way to do it using science, which requires working within the laws of physics.

      So far, no fool cell fanboy has responded to that challenge.

        1. Pushmi-Pullyu says:

          Your link has nothing to do with using hydrogen as the fuel for fuel cells. That’s just using excess energy to generate hydrogen at a significant loss of energy — even if you take their website’s claims at face value, the average one-way energy loss is nearly 30% — and the hydrogen is injected into natural gas lines, where it can displace up to a maximum of 15% of the natural gas.

          You can’t run a fuel cell on a mixture of 85% natural gas and 15% hydrogen!

          Compare to using batteries for energy storage, where a two-way or “round trip” energy efficiency of about 85% can be achieved.

          Try again, “Electron”.

    2. Ambulator says:

      “… with hydrogen being the most abundant element in the universe, and on earth, …”

      In the Universe, yes, on Earth, no.

  4. Get Real says:

    Hey, another new username here to post pro-H2 propaganda. I wonder which of the existing usernames that are universally anti-Tesla/pro H2 this one is?

    Your fantasy doesn’t work electron because of the eternally unchanging laws of physics.

    See here:

    The simple fact of the matter is at best, H2 is 1/3 the efficiency of using renewable electricity into a PEV’s battery and therefore it is uneconomical and wholly impractical.

  5. electron says:

    lol…get real did you even read my post…..batteries are better now because of real hard science r&d improvements, and in the short term is the better solution….but for you to deny hydrogen as a solution in the future is rather short sided…..for the long game hydrogen as fuel has to be the major player or we as a species won’t be able to exist, atleast at current population growth projections….I am not anti tesla in anyway, I love them, just too expensive right now…..sea level rise is a foregone conclusion, we are not going to do enough soon enough to stop it…..I am just being realistic, don’t you think it would be wise to try and find a good use for all that rising water ? what do you think made all that electricity for your battery anyway (hydrogen fusion) ?? more energy efficient hydrolysis and de-salination will be critical for a large part of the world population…..solar and wind won’t power the world, or a space x rocket…..we are gonna need both and the sooner the better……

    1. Pushmi-Pullyu says:

      electron said:

      “…but for you to deny hydrogen as a solution in the future is rather short sided…”

      You mean “short sighted”, but no, that’s just being realistic. There are a number of fuels that would make practical everyday fuels if synthesized using renewable energy. Hydrogen simply ain’t one of ’em; it wastes too much energy and is too difficult (and expensive!) to work with.

      Changing hydrogen to be a practical fuel would require changing the basic physical properties of an element. Now, is it that you’re not scientifically literate enough to know that’s impossible, or is it that you don’t care because you’re just here to shill for Big Oil?

      Big Oil loves to promote hydrogen fuel, because they know it’s an unworkable and completely impractical technology, one which could never compete with using petroleum as the primary source of energy for transportation.

      “…don’t you think it would be wise to try and find a good use for all that rising water ?”

      Finding hydrogen ain’t the problem. It’s the most abundant element in the universe. Using hydrogen as a practical everyday fuel… that is the problem which is physically impossible to solve.

    2. ItsNotAboutTheMoney says:

      That’s really not true.

      Humans are currently trying hard to figure out the least worst way of working around the variability of renewable energy.

      The advances in renewable generation and battery technology have been great enough that we appear to be on a path of using renewable electricity, batteries and electric vehicles where possible, but we need to fill in the gaps.

      Hydrogen is a _possible_ way to fill the gap. But there’s also ethanol and, in particular, methane.

      Methanation would add another step to the synthesis and therefore introduce more cost and losses, but methane is a good candidate because we already use it and store it in very large quantities and it’s much easier to store and use than hydrogen.

      Also, we can and do generate quite a large amount of biomethane anyway, and even in a fully renewable system, methane and other hydrocarbon use wouldn’t go away, since they are very useful chemicals.

      1. Pushmi-Pullyu says:

        Yes, exactly. Methane is much easier to work with. Unlike compressed hydrogen, methane does not require specialized high-pressure pumps or storage; does not require special seals for storage tanks, pipes, and fittings; does not embrittle metals in those same items; and can be transported via existing natural gas pipelines.

        The initial generation of synthetic methane may or may not be as cheap as generating hydrogen, but it’s much cheaper by the time it’s actually dispensed into a vehicle’s gas tank.

        In short, synthetic methane is a candidate for a renewable, practical everyday transportation fuel, which hydrogen will never be. That’s why you’ll never see Big Oil shills promoting the idea of using renewable synthetic methane rather than compressed hydrogen.

  6. electron says:

    pay close attention around the 30 minute mark !!

    http://ihavenotv.com/what-is-our-future

    1. SJC says:

      Some can not pay attention for 30 seconds.

      1. electron says:

        lol maybe i should say fast forward to the 30 min mark, but watch a bit longer than 30 seconds it really is worth a watch….

        1. Pushmi-Pullyu says:

          Why? Did they figure out how to repeal the laws of physics?

          A lot of perpetual motion believers want us to watch a video which “proves” their claims, too. No mere video can possibly refute the Laws of Thermodynamics.

          1. SJC says:

            P-P
            Mellow out, this is not an insult contest.

  7. Pushmi-Pullyu says:

    Toyota is trying to spin the fact that the company is gonna have to play catch-up in the EV revolution, as if their R&D efforts are somehow something different or better than what everybody else has already been doing for years.

    Reality check: Battery tech R&D groups have been using expert system software (often given the misleading label “A.I.”) to search for useful advances in chemistry and the internal structure of batteries, for several years now.

    Linked below is an article from Siemens which references development of such an expert system in 2009, and for all I know, other companies may have developed similar software even earlier.

    http://mdx2.plm.automation.siemens.com/sites/default/files/brochure/pdf/Battery%20brochure%202014%20final%20pages%20low%20res.pdf

    * * * * *

    Dear Toyota:

    Nice to see you finally stopped trying to swim upstream; stopped denying BEVs are the future of automotive technology. Now take your lumps, quit trying to pretend you’re not trailing the pack in EV development, and if you must continue to make “fool cell” cars to capture the ridiculous amount of money the Federal and Prefecture governments in Japan offer for that unworkable and wasteful technology, then at least stop trying to convince us that there’s any future in that tech.

  8. brendan says:

    It has always surprised me how Toyota continues down the Hydrogen path when EV’s have shown to be a far more superior solution for most drivers. I can see applications for large transportation such as trucks etc, but for standard commuter vehicles i dont get it. Come on Toyota you could make some great EV”S!
    http://www.evse.com.au

  9. Sparkinator says:

    Just when I thought that Toyota was admitting their blunder into FCV, then they say well short term we’ll do this battery thing, but really FCV is where we’ll be in 30 years. Is it pride, or some kind of mental defect?

    Fool cells are for those who own stock in Big Oil – nothing else. We do not have to burn something to move our vehicles down the road.

  10. Ocean Railroader says:

    I could see using a highly advanced computer would be good for something like battery tech.

    Such as if you were to give the computer hundreds if not thousands of battery chemistry mixes or formulas for battery mixes and it runs simulations on it. It could save a lot of time and money unless battery chemistry is one of those things were it might behave differently in real life then on a computer.

    1. Pushmi-Pullyu says:

      Certainly any given chemistry or cell configuration might work differently in real life than it does in a computer simulation; the simulation is only as good as the knowledge of the people who wrote the program.

      The advantage is that expert system software can be used to weed out most of the almost infinite number of possible combinations which clearly won’t be as good as the battery chemistry and battery tech we’re already using. They can be used to find certain chemical combinations or cell configurations which are worthy of real-world testing.

Leave a Reply