• An up-and-comer in EV batteries…
  • Interested in lower car insurance?
  • An emerging trend in computing for 2023…

Dear Reader,

It started with a small wobble, just enough to move my chair and cause me to look up at one of my colleagues and smile.

I was in my office on the 24th floor of a 32-story skyscraper in Tokyo. What happened next was unbelievable… 

That little wobble turned into a powerful vertical thrust coming from the ground. I could feel the entire building being lifted several inches, drop, and then lift again. For anyone who has experienced earthquakes before, that’s a very bad sign of what’s to come.

Moments later came a slap, a wave of raw power unlike anything I’d felt before, one traveling horizontally that caused the entire building to emit the sound of a large “crack.” Drywall buckled and shattered in the stairwells, and filing cabinets were banging against each other everywhere.

The sounds coming from the building as it was stressed to its limits was eerie…

It was March 11, 2011, the day of the Great East Japan Earthquake. A day I’ll never forget.

In the days that followed the massive earthquake, and meltdowns of three nuclear reactors in Fukushima, most ex-pats working in Japan took their families and left the country. I stayed.

How could I possibly ask my team to “keep things running” and manage the crisis while I shuttled off to Hawaii to manage from afar? No way. I sent my family south to Fukushima – as my first son was still an infant – and I stayed in Tokyo. I bought radiation meters for my offices, as well as three satellite phones, and buckled down for the crisis that I knew would ensue.

I told my team that if they were uncomfortable staying in Tokyo, they could take their families anywhere they’d like and work remotely, and I’d pay for it.

They all stayed. And we managed the crisis together. It’s something I’ll never forget.

I was, and still am, deeply grateful for the teamwork. That event, and the months that followed, was one of the most formative life experiences I’ll ever have. I’m a far better person, and executive, for it.

The earthquake, and the meltdowns, not surprisingly had an outsized impact on Japan’s energy policy. All of Japan’s nuclear fission reactors were shut down. Cities were dark at night to conserve power. Escalators were turned off. 

When summer came, air conditioning was set to 28 degrees Celsius (82.4 °F) for the same reason. We could feel the effects everywhere.

In the absence of nuclear fission power plants, the country had no choice but to turn to coal and natural gas just to keep the economy running. It went from carbon emission-free power production to fossil fuels powering the entire country seemingly overnight.

And that energy policy has persisted for the last decade… Until now. Here’s what has happened:

Since the beginning of 2021, the Yen has weakened by 40% against the U.S. dollar. We haven’t seen a Yen this weak since 1990. And since Japan has to import its fossil fuels, the burden has simply become too high to sustain.

The country has hit a tipping point, and a dramatic shift in energy policy is underway. Japan is now planning to turn back on seven more nuclear fission reactors, which would bring online a total of 17 (of 33) operable nuclear fission power plants.

The reality is that the meltdowns in Fukushima weren’t the failure of nuclear fission technology… they were the failure of management and bureaucracy. A post-event analysis from the Carnegie Endowment for International Peace summarized how simple it would have been to prevent the meltdowns:

  • Protecting emergency power supplies, including diesel generators and batteries, by moving them to higher ground or by placing them in watertight bunkers

  • Establishing watertight connections between emergency power supplies and key safety systems

  • Enhancing the protection of seawater pumps and/or constructing a backup means to dissipate heat

These simple and inexpensive measures would have avoided the meltdowns. Nuclear fission technology is safe, as long as best practices – which are well understood in the industry – are employed.

A friend of mine who lives in California wrote to me this week about the importance of nuclear fission as a source of clean energy. His background is relevant as he used to work on British naval nuclear submarines back in the ’80s. After that, he consulted in the power generation industry because of his experience.

He reflected on how far the nuclear industry has progressed in terms of building and running nuclear power stations – and equally important, the safe decommissioning and storage of spent nuclear fuel.

He’s right of course, the industry is nothing like it was back in the late ’70s, around the time of the Three Mile Island partial meltdown in Pennsylvania. The technology and processes for safe operation and storage have transformed since then.

But in the U.S., the availability and safety of the technology won’t be enough to make it a reality. There is too much irrational political resistance from too small a part of the population.

Yesterday, California’s Independent System Operator declared a stage 2 “energy emergency alert,” warning that California residents should prepare for rolling blackouts. Californians are also being asked to turn up their thermostats and reduce electricity usage to help with the crisis.

But this won’t be enough to cause a tipping point in energy policy. It’s temporary. And the availability of electricity isn’t a problem at all in almost every state. The only national “pain” is the rising price of electricity.

That’s not enough to overcome the political opposition. It would take a lot worse… At least as much pain as is being felt in Europe right now, or what’s happening in Japan.

But the reality is that the U.S. has the ability to become energy independent again without new nuclear fission reactors. It has the natural gas reserves if it wants them.

But I’d much prefer a 100% clean energy grid, without any use of fossil fuels – the kind of power that can run factories, and supply offices and homes with clean energy 24/7 all year long, rain or shine. 

That means that there is only one politically acceptable solution: nuclear fusion power generation. And the sooner we employ it, the better off our world will be.

Solid-state batteries are about to become a reality…

Next-generation electric vehicle (EV) battery company Factorial Energy is making a big move…

I’ve been monitoring this company for quite a while now. Factorial Energy is one of the key companies that has been working on solid-state batteries for EVs.

But the company has been very tight-lipped about its progress. It has intentionally stayed under the radar… Until now.

Factorial Energy just announced that it is building a $45 million production facility in Massachusetts. The plant will mass-produce solid-state batteries for EVs.

And here’s the kicker – Factorial Energy expects it will be operational early next year. That’s what makes this so interesting.

No company has been able to mass-produce solid-state batteries. The technology, at the scale required for an EV, has still been under development. The technology is real, but there are still many challenges to solve with such a large battery. And automotive safety requirements are quite stringent, making it an even larger challenge.

Most companies working on solid-state technology have focused on producing smaller prototypes as a proof of concept. Some are using consumer electronics applications as a stepping stone before tackling EV batteries. 

Factorial Energy had been quiet about its progress, so the assumption was that it was a ways away from production. But this latest announcement signals that Factorial may be farther along than originally thought.

Here’s an image of Factorial’s prototype EV battery:

Factorial Energy’s Solid-State Battery

Source: Factorial Energy

As we can see, Factorial is making full-sized batteries. Of course, it will look different once packaged for an EV, but the battery pictured above has the storage capacity to support EV applications. And the fact that Factorial is pouring $45 million into a production facility signals that management believes the company can commercialize its tech right now.

And if we look at Factorial’s early-stage backers, the picture becomes even more clear.

Factorial Energy has taken on over $250 million in investment to date. That includes a $200 million funding round back in January.

It could have taken money from just about anywhere given how hot the EV battery market is right now. But Factorial has been strategic about who it has taken investment from. The list includes Mercedes Benz, Hyundai Motor, Kia Motors, and Stellantis (the parent company of multiple auto brands, including Fiat/Chrysler).

Clearly, these top-tier carmakers are excited about Factorial’s solid-state batteries. They are likely to be the first customers when Factorial’s production goes live next year. So it was logical to take investment dollars from these customers. They will be naturally incentivized to see Factorial be successful.

So this is an incredibly exciting development.

If Factorial can, in fact, begin mass-producing solid-state batteries next year, it will be the first battery company to go to market with solid-state batteries for EVs. That would be a huge accomplishment and an incredible competitive advantage.

Seeing is believing though. It’s great to see the investment, so now it’s time to see the actual results and performance of the manufactured batteries. I’ll be looking forward to seeing how these perform.

The future of insurance is here – built-in incentives for better driving…

A small, private company called Zendrive just made a big announcement. This is something that will upend the commercial insurance industry… And very few saw it coming right now.

Zendrive has developed software that uses the sensors that are already inside of a smartphone to measure driver behavior. This allows Zendrive to create a safety profile for drivers over time.

Do they drive too fast? Do they brake too hard? Do they take turns too sharp? Are they generally aggressive? Or do they practice defensive driving techniques?

Initially, Zendrive’s goal was to collect all this data to sell it to insurance companies. The idea was that armed with this data, safe drivers should have lower-cost insurance. And drivers who exhibit risky driving behaviors should naturally pay higher prices.

This makes perfect sense. This kind of data can be used to incentivize safe driving, which ultimately leads to lower car insurance. Even better, the net result would help reduce the number of car accidents substantially over time.

And it turns out this service has been a huge hit in the marketplace.

Zendrive now has more than 50 million drivers in the U.S. using its tech. And the company has collected more than 200 billion miles of data from hundreds of millions of drivers around the world.

This is by far the most expansive safety driving data set ever assembled. And as regular readers know, machine learning (ML) algorithms can extract a ton of insights from a large, detailed data set like this.

And the big news is that Zendrive plans to capitalize on this great success. The company just announced that it is spinning out its own commercial fleet insurance firm, called Fairmatic.

Fairmatic will leverage Zendrive’s data to provide ultra-competitive insurance for commercial fleets. It will customize commercial policies to a degree that’s never been possible before.

This is the future of insurance. It’s here now.

The days of standardized insurance costs where safe drivers subsidize risky drivers are over. Fairmatic is creating a marketplace where costs are directly linked to driver safety.

And what’s incredible about Zendrive’s approach is that it didn’t require any expensive hardware to be installed in every car, and it doesn’t require any kind of complex agreements with automotive manufacturers. The solution simply utilizes technology that already exists inside every existing smartphone.

This is an inflection point for the car insurance and commercial fleet insurance industries. In a few years’ time, this will be standard practice in the industry, and it all started with an early stage private company.

One of the hottest tech trends for 2023…

We’ll wrap up today by analyzing the seeds of what will become one of the hottest tech trends of 2023: artificial intelligence (AI) running at the edge of networks.

As a reminder, data is usually transmitted back to a massive data center full of processors. There, thousands of servers run AI software and process all of the data. Once complete, the results are sent back to the “edge” of the network where the results can be employed.

Yet advancements in recent years are changing where AI software runs. More and more, AI can now function out in the field and perform tasks right there on the spot. No need for the massive data center.

And an early-stage company called SiMa.ai just made a major breakthrough that will power this trend forward. SiMa developed a highly integrated platform that enables machine learning (ML) to run on a single chip.

SiMa’s product is called a machine learning system on a chip (MLSoC). Pretty straightforward. Here it is:

SiMa’s MLSoC

Source: SiMa

It may not look like much, but this device will power all kinds of “intelligent” devices. That’s what AI running at the edge of networks is about.

For example, think about a security camera powered by AI. It could detect a certain kind of motion, and the AI would decide what action needs to be taken. Maybe that’s recording the scene… Maybe it’s zooming in… Maybe it’s alerting the owner or law enforcement authorities.

Whatever the decision is, the AI will come to it by “inferring” the best course of action. That’s what SiMa’s MLSoC will enable.

The number of applications for this kind of technology is limited only by our imagination. We’ll see a proliferation of smart devices operating at the edge of networks using this kind of technology next year.

And the beauty of SiMa’s system is that it incorporates both the hardware and the software.

SiMa built the MLSoC hardware that we see in the image above. And it developed the software platform upon which any kind of AI or ML can run. We can think of the software as the operating system. It allows anyone to run custom applications on top of the system.

This is exactly what NVIDIA did to become a semiconductor powerhouse. NVIDIA developed the hardware – its graphical processing units (GPUs). Then it developed the software that developers could use to create custom applications powered by the GPUs.

So SiMa is taking a page out of NVIDIA’s playbook here. It’s a great strategy.

And best of all, the company plans to begin mass production of its MLSoC product in the first quarter of next year. They are past the prototype stage and gearing up for commercialization. We’re just a couple of months away.

This is something that will achieve rapid adoption very quickly. Lots of companies will tailor SiMa’s technology for their own applications. So this is yet another great early stage company we need to track.

I expect that we’ll be exploring this topic a lot next year in The Bleeding Edge. And the focus will be much more on how this kind of technology is being used in the field, not in a laboratory. It’s going to be fun to share real-world applications that bring this kind of technology to life.


Jeff Brown
Editor, The Bleeding Edge

Like what you’re reading? Send your thoughts to [email protected].