Well, the holidays have come and gone once again, but if you've forgotten to get that
special someone in your life a gift, well, Mint Mobile is extending their holiday offer
of half-off unlimited wireless, so here's the idea.
You get it now.
You call it an early present for next year.
What do you have to lose?
Give it a try at mintmobile.com slash switch.
Father.
Or should I call you Steve?
Since it's a new year, maybe we should start cooking fresh, with fewer smoke alarms this
time.
Your cooking's improving, but I can see the stress in your eyes.
Anna's mom used this Blue Apron Assemble and Bake thing during our play date, pre-chopped
ingredients, over 40 grams of protein, and just one pan to clean up.
She just tossed it in the oven and boom, plating like she's got followers.
So go to BlueApron.com Steve, make 2026 the year you win dinner.
Blue Apron.
Get 50% off your first two orders, plus free shipping with code LISTEN50, terms and conditions
apply.
Visit BlueApron.com slash terms for more.
Dear winter, Toyota can't get enough of you, because Toyota's got 25 vehicles with available
all-wheel drive and four-wheel drive, and that's more than any other auto brand.
From the versatile RAV4 to the svelte crown, the sleek Camry all-wheel drive, the Corolla
hybrid all-wheel drive, the rugged Tacoma, the tenacious Tundra, and the spacious Grand
Highlander with all-wheel drive to keep you and your safe.
Hey, you bring the action, we'll bring the traction.
Toyota, let's go places.
Based on manufacturer's website says a 10, 20, 25.
The industry is a little bit drawn in this marketing game on competing on this kilowatt
hour.
These numbers are only meaningful if you actually look at it in combination to other aspects
of the battery performance.
How long the battery lasts?
When it's aged, can you still sustain the same performance?
And probably like how much energy you utilize across your battery pack?
Are you able to reduce your mass or volume if you are not utilizing all of it?
So this kind of question needs to be considered in conjunction with this single-number digital
charging power.
Hi, everyone, welcome to the CarChat podcast.
I'm Sam Mores and today I have Jingyi Chen.
Hello.
Hi.
Hello.
Can you tell the listeners a sort of short summary of who you are and what you do?
My name is Jingyi, I'm the head of battery at Breeze Battery Technologies.
We are a business working on battery management software and looking after the whole two-channel
of how we can provide software as a channel to improve battery performance.
So I'm actually a material scientist by training, nothing to do with software to start my journey.
But I had a PhD in electrical engineering in field cells, but I've always wanted to work
towards electrifications because I just believe Electron is the new gasoline.
So kind of it.
Maybe I go a little bit about what we do at Breeze.
So on the day-to-day, we work very closely with our OEM clients.
They are car manufacturers, laptop manufacturers.
So we look after the pinpoints they have in either they feel the battery run up too quickly
or they have to replace battery too frequently.
So these kind of pinpoints, but we look more into sort of chemistry-related characteristics
of batteries like energy densities, for example.
And we try to understand the physics behind it and offer software solution to help them
improve their battery performance.
And we do it for different applications.
Sometimes the laptops may want less heat from the battery.
Sometimes the vehicle wants sort of faster charging.
So slightly different application, but all looking into how we can improve battery
by physically understanding it better and utilizing those physical understanding.
Interesting.
Okay.
And if you're looking at laptops and car batteries, are they in a similar state of technology?
I know laptops have probably been pushed for a long time, but in terms of do they scale
nicely from one to the other?
So obviously when you look at them, one is pretty tiny, the other is like big chunk.
But fundamentally, they are all a lithium-ion battery technology, so which means they in
the electrochemical sense, they all work fundamentally the same.
The mechanism is what we call an intercalation.
So you have lithium-ions being shuffled around the cathode and anode and you have electrolyte
to wet it around so the ions that can actually migrate.
So that's fundamentally if you view them layer by layer to sort of look at them atomically
in a way.
They are the same, they're just being built in different sizes.
Right.
And so you guys, you're looking at the base, what are the base technologies we're looking
at at the moment are lithium-ion.
But we have in car batteries, they're not all the same at the moment are they?
So in car batteries, they are all belonging to kind of a lithium family.
They are emerging sodium technologies just because sodium is a more accessible chemical
than lithium.
But all of the car battery nowadays, EVs or plug-ins, they all belong to this lithium-ion
family.
Some people might have explore like solid state battery, lithium metal battery.
But those technologies out there are all sort of lithium based.
Okay.
So it's a big family and then they're just working different stuff in between.
And then why did Breathe come at this from, you said you're looking at, you know, the
base physics, et cetera, but then you're coming at it from a software point of view.
What can software, how can software manage this stuff?
Because I presume every, I don't know, you'll be able to tell me when a manufacturer comes
in, they probably want, I want, if it's a car, more range, faster charging, more durability
and to operate in every single temperature range possible, that they just want more,
more, more, more.
How do you, you know, where do you come in?
Yeah.
So I either say software is only one part of the ecosystem.
There are many efforts that both the OEMs or the cell manufacturers, battery manufacturers
do in terms of design or system integrations.
So I, the way, the reason why software has a big part to play is I think fundamentally
if you look at batteries, it's an electrochemical devices.
So if you are able to, so it's, it's more complex than the traditional internal combustion
engine where it was more thermal and heat that people work on.
But now we work on chemical reactions inside it.
So where we can look into is those chemical reactions are not linear.
So the study of it depends on sort of analytically describing those nonlinear relationships happening
because of these reactions happening.
So this is where what we focus on the physics-based software tool chain, we try to understand
the physics and the, then the reactions happening inside the software and describing it
by using software as a platform to understand the battery either offline for a more complex study
or online to do real-time control.
So this is where I feel software has a big part to play in EVs in comparison to the old
internal combustion engine because of a complexity in there.
The other way to look around it is because the navigation of an EV is heavily relying on
software and control technologies.
So it's less mechanical than combustion engine.
So that actually provides a very good software ecosystem for sort of a more complex software
like us to be able to integrate it into that system just because of also the platform is
being upgraded to be more sort of software elegant.
Right. So are you saying that the base, the way a battery works is actually super complicated
like in terms of how it outputs all the stuff.
It's not just, we don't 100% understand and we're trying to learn is that what you guys
are working on and then presumably the management of that.
Yeah. So this is the thinking at the breeze that we believe understanding the physics is
a pathway forward.
So essentially if we look at a battery in an EV, so if we want to probably want to understand
a little bit of what happened to your car battery, you probably want to measure the temperature,
measure the voltage, look at the current you draw while driving.
Basically like voltage, current and temperature are what you can put a sensor on
and actually get reading.
But inside the battery, while this is happening, it's actually the reactions I mentioned like
ions shuffle around.
You kind of want the ions to inter clate into your electro materials.
You don't want it to sort of form something called plating that we have mentioned.
So that you kind of want the ions to be usable, but not some other reactions that render it just
bad.
So that's the understanding sort of the physics provide.
And we have to get from a direct management of voltage, current and temperature to that
physical understanding.
We have to go through some kind of modeling or process or a data driven process to
get us the actual understanding of kind of the black box inside the batteries.
Okay.
And then those are the only three bits of information.
And do you get those at one point, just output of battery?
Or are people starting to have sensors sort of in the chain somewhere?
So ideally the more depending on the geometry of the battery, you probably have a bit of
localized effect.
So ideally the more you sense, the better information you get.
So you can capture some local effect because nowadays the battery in EVs are being made
bigger and bigger.
So if you look at that, there's this chunk of, yeah.
So there are some local effects.
If in theory, if you are to monitor it everywhere, then that's the best solution.
Practically like putting each sensor adds you a little bit of a cost inside it.
So you have to be quite strategic in where to place it and how many sensors you place it.
And then the quality of sensors you choose is mainly around the cost.
So what we try to do is we also look at the client's sensing profile.
And we try to make our software to be able to handle some of it, utilize some of the
informations and in where we feel that there is some additional care we need to be taken
care of because of a limited sensor.
We provide suggestions or solutions to put in some margins in how we control it.
So we make sure that we always utilize most of the information we get and optimize how
we get out of the battery.
But unavoidably, in some places we have to be quite careful in how we treat off.
So safety is always the number one thing we do not want to touch.
So we probably want to navigate away from and put some margin in the safety.
And so you've got all this information, you're working out how it works.
And then what is your software actually doing?
What's it controlling and what's it trying to do?
So we actually have like two ends of the software suite that we offer.
So on the control end, we have this embedded software called Breeze Charge.
And what it does is it's reading those sensor information, a voltage current and temperature.
And it tries to, it calculates in real time, what is the optimal charging current
based on the reactions happening inside the battery.
And the most famous one that we try to navigate away is called lithium plating.
So basically, the lithium ions just become that lithium metal if this reaction happens.
So we navigate away from lithium plating.
And by doing that, we can sort of calculate an optimal current, typically the highest current
that the vehicle can charge at each instance of time.
And that will enable either a fast charge experience
without any compromise in the durability of the vehicle.
Okay. And then, and what might that like, if you looked at, I know from reading magazines
and whatever, a sort of typical charge curve, it was to start fast.
I don't know if it's zero, I guess it's also going to be slow, but let's say from 10%,
it's going to be really fast and then it's going to decline over time.
Is that because this software is managing but isn't perfect, for example, whereas if you're
like, you're managing it perfectly and you're keeping it within the optimum stuff,
is it kind of like, is it almost going like up and up, down, up, down, up, down, up, down,
just like max for a tiny bit, then it has to throttle it back for a millisecond.
What does it look like when you're optimizing it?
Yeah. So actually it has gone through a quick, sort of long and learning journey in the battery
industry. So at the moment, everyone kind of know to stop fast and drop in terms of power.
But when it first started with sort of like early understanding, it was kind of kept at
constant power or constant current. And as the understanding evolved and we realize
the best way forward is to stop fast and slow down later. So that's actually an improvement
in learning. It's not a lack of learning. The reason why it needs to do that is
mainly because when your battery is empty, you're quite far away from plating
and any kind of your sort of negative electrode. So you can pump in a high current.
So as if you can turn the tap on and then you don't worry too much about overflow.
And but as it being full, you need to kind of
you're closer to sort of these trigger points of the side reactions. So you don't want to
keep your tap so high. You're going to close the tap a little bit so it doesn't overflow.
So that's roughly the thinking behind it. And what causes the plating?
So it's kind of a competing mechanism inside of it. So fundamentally, it's a thermodynamic
threshold that if your negative electrode potential reaches zero volt, and then the
condition is more favored towards plating rather than in intercalation. So that's a very
theoretical value. But in reality, if you have a magic probe and put it on every little point
on your electrode, you probably can measure it. But in reality, it's always done through estimation
towards how your electrode potentials states are at each instances of time.
Okay, so this is how your software, the better it gets, the better your understanding gets.
You can charge a battery faster because you know you're tending towards that limit.
They're like the real limit. And you want to maximize that.
And when we are away from the limit, I actually want to turn the tap a lot more powerfully here.
And what's this with real world meant in terms of difference in charging times? Because I presume
let's just say same battery five years ago, whatever it is, versus same battery now charging.
Yeah. So I think when I first entered into the industry, probably five years ago in the
lithium ion and EV space, people were talking about roughly half hour, 14 minutes charging time as
the target for fast charging. But I think today we are actually talking about 10 to 20 minutes
charging. So as the understanding evolved, we can actually reduce it at the industry as a whole,
have already reduced it by roughly three times since. And that's same, same charger, same battery?
Just software. Yeah, I would say software alone wouldn't have done this like so significantly.
So I do not want to deny the efforts by other people working on this field. So there are also
chemistry and manufacturing improvements on how we can stabilize those materials to avoid plating
in there. So it was actually a combined effort in both cell design and software I would say
that plays the biggest role. But what do you reckon just software? And like how far have you
since you guys started to where you're at now? And then I guess the lead on question is
I guess the gains are going to get harder, but how close, you know, you know, how far are you away
from the perfect charge in terms of that? I would say like there is more we can do on the
software side looking beyond the control, sort of the control sense, we typically if we are to
sort of bring to the market, embedded control software, we typically enter the production
cycles of EV pretty late. So pretty like pretty two or three years before the production is the
normal entry point. But before that, there is typically a five to 10 years efforts in sourcing
the right battery and finding the correct design of the vehicle pack. I think that's where a lot
of software efforts can also help in that if we can understand sort of offline simulations or
how we can optimize the design when we make the actual batteries, when we source the materials
and how to synthesize the material. So that is, I guess, a modeling efforts that we can put more
into. Yeah. Okay, so a bit like I'm redoing a house at the moment. And everyone that's doing
the design wants to be involved before we've started building the house. Because you're like,
if we do it now, if we plan everything properly before we actually build it,
you can do a lot more. I guess that's the same. You don't want to bolt on your software
when someone's finished a car. You can go, well, we can help, but it would be way better
if we were in early. Yes, exactly. Yeah. Yeah. I think people are already doing that. I think as
the industry mature, there will be a big area that everyone wants to tackle. And if we really want to
go down into probably like five minutes charging. Yeah. And so in terms of this state of where we're
at now in the industry, I know you've seen a lot of this, what are the sort of big innovations
that you've seen that you think either implemented at the moment or coming or on the horizon that
you think might make a big difference or help in this space? What I felt is the big innovation is
actually the synergy between an offline simulation tool and the online sort of management tool and
how they can combine to sort of bring up a more enhanced effect of how people can use it. This
is where I've seen like, we touch on the sensing point, actually offline simulation nowadays can
give everyone a pretty accurate result of the distribution of current voltage and temperature
when you have this gigantic brick of battery. So that knowledge can be utilized onto real-time
control. And I believe that the whole industries are already starting to do that. And the learning
from the online control can be then fed back to the offline simulation when you're trying to
design a system. So this is where I just say the last couple of months the realization I had
and this is an area that has great potential. Cool. And then so with the current like physical
battery technology, I know some companies have made battery packs that you can charge
if I think of like, was it there like rematch ages ago, they could charge it 500 kilowatts.
I think BYD have come out with a battery pack, they're claiming they can try to charge it,
what is it, a megawatt? Is that a thousand kilowatts? With those systems, how does in real
world, I think one of the things when I was talking to the rematch people is they're like,
yeah, we can, our max charging is 500, but there aren't really any 500 kilowatt chargers,
like there are some, but real world, you're not going to get many. But if you've built a structure
and system that can charge it a megawatt, does that mean you can plug into a 250 and just bang
250 all the way along? Or like what's the charging curve when you've got that capability
on a sort of more standard charger? So you're absolutely right. That is probably one or two
500 kilowatt chargers out there in the world. In actually, in my very humble opinion, I think
the industry is a little bit drawn in this marketing game on competing on this kilowatt hour.
I always told my engineers, these numbers are only meaningful if you actually look
it in combination to other aspects of the battery performance. How long the battery lasts, how
when it's aged, can you still sustain the same performance? And probably like how much energy
dense, how much energy you utilize across your battery pack? Are you able to reduce your
mass or volume if you are not utilizing all of it? So this kind of question needs to be considered
in conjunction with this single number digit charging power. So this is where I felt actually
the industry is in a state of thinking about this problem and maybe one day start to move away from
that one number because I think at this stage to your very correct realization, it's not
providing actual use of benefit if there's no charger out there for you to write on those numbers.
So but if we are moving resources away from this single number game, we can look across
more benefits that we can provide to the driver than the end user.
Yeah, it definitely feels like that to me. You get the more I learn about these technologies,
the one that efficiency is probably the number one for me, like real world actual
sort of efficiency of these vehicles. But charging power, I think to some people feels a little bit
like horsepower. It's like it's not the same thing. But it's you know, it's like, oh, my car can charge
this fast. And you're like, yeah, so what like, does it in the real world, is that beneficial to you
or not? I've seen some technologies, I know some batteries, I believe Porsche have it where they
they have an 800 volt battery pack or system, but they can split the battery pack into two
400 volt systems and charge, I don't know whether they charge in parallel or they charge sequentially
or whatever to help charge faster. There's these sorts of things like breaking a pack up into
different sizes. Do you know a little bit about this? So I would say that pretty touch on the topic
called self balancing. I don't exactly know the Porsche technology in particular. But to break
things up, basically, if you would like to have sort of two parallel design, you probably want to
balance those two components. So you don't age one part more than the other. So this is also
again, another area that we at Breeze pre don't look too much into but other people in the industry
have look into the balancing topic. Yeah, I was just wondering if there's you know, what are the
the real world things that people people doing to sort of help help with all this sort of stuff
and keep it keep the cars cars going and whatever do you aim versus like where we are now and I know
from the cars that sort of come out, we're getting a couple of percent a year maybe kind of efficiency
gains over in the whole system and it just keeps rolling and rolling and rolling and
that's meant quite a lot over the last 10 years. But do you see that in these sort of technologies
just rolling like that slowly over time or do you think there's going to be there will be
solid state will just be a step change or do you think that's kind of like a I think that's off
that may not really make much difference. Yeah, so I think the rolling will continue in a in a sense
like people will try to make incremental changes as we developed and then have better understanding
of the system. What I do see a breakthrough basically it can happen in in two fronts.
One is if we have a breakthrough in the materials our discovery I think probably like
if we look at them even the solid state market then if we compare to the lithium ion battery,
they kind of have their pros and cons with each other so no one is no material is like superior
to like the point of it. So it's also it's always a gain of balancing and the trade-off
in that front but if we can I don't know how many years it will take or some brilliant scientists
may have like a spark one day to have a material that is still so superior and it becomes a leading one
and sort of bring this like innovation breakthrough in there. The other thing I can see is particularly
around understanding the physics and combining it with sort of data we get as we accumulate more
vehicle data in the field or more testing data as we developed particularly when we are able to
test more aged batteries so that understanding will actually enhance how we complete the overall
understanding of battery how it behaves when it's new and fresh and intact in a sense but
when it's being used for maybe five to ten years how the electrochemical behaviors have
changed over time but once we complete the full pictures over life and just as humans like
proud to an old man then we can have a full understanding of a battery and that will bring
the real breakthrough here. And presumably has the management software gets better
you're like well whether it's charging or just general usage the batteries we've got that are
10 years old now had 10 year old management stuff on them so you're kind of like the stuff you're
putting in now you're like well we need you know we need five years 10 years or well I guess I guess
batteries presumably get tested in on rigs and stuff. Yeah so on read they're also like these
years of testing but the difficult parts in battery is if you manage them differently they will age
differently yeah so that's the tricky part where we need potentially some exponential growth of
data points and understanding in that yeah so so that will come from like more testing and sort of
and plowing plunging through the data is not a difficult it's not a easy thing if you are just
taking a very sort of mathematical data driven perspective so I do believe like even delving
making actual use of the data needs to be combined with physical understanding of the
reactions and the happens that has happened and potentially will happen inside the battery.
And if you look at cars out there at the moment are there some brands that seem to be doing it
better or seem to have done it better and their batteries hold up better than others or is it
is it car brand or is it manufacturer of battery or a blend of both?
Pretty I can start with the car brand I wouldn't say one is doing better than the others I think
the actual design of the car reflects the branding messages of each of them some may
want to be go like some may prioritize your ability and for like a common commuter use
some may prioritize a hypercar performance so I would say the target audience and the
application is different and that renders them to be perceived differently in a sense.
In terms of how the manufacturers evolve I think that at the moment there is a trend
for everyone in the world to try to diversify the supply chain at the moment it's very concentrated
in the East Asian area but there undeniably there is a lot of learning and mistake people made
and sort of grown into this like large manufacturing hubs so I would say in that so there are some
joint like manufacturers like CATL that they have years of learning and they're probably like
knows quite a lot about like everything but there are some niche suppliers who has niche
understandings on certain performance if you want to push forwards you have some target might push
forwards so they are kind of scattered around this is my opinion of how the industry are at this
stage I would say what I see as a future trend is closer collaborations between the manufacturers
and OEM in terms of the OEM pre like car manufacturer will try to feed the battery
manufacturer supplier sort of their demands their pain points their use cases earlier in the stage
and try to engage more in the design phase of the battery and the sale manufacturer will then
need to listen to more of the end user impact rather than just selling to those business.
Okay and then as a sort of final user car or EV user question one thing that always comes up
when people talk about is how often you fast charge your car is that
is that just as like a general bit of advice to users now is should you not fast charge your car
if you can avoid it should you only charge it to a certain percentage what are your thoughts on those
things in in the real world of people's actual usage. Yeah so I heard this worry a little bit
if you don't try not to fast charge your car as often as you need but actually
if you there are some reports out there either like on YouTube or people try to tear down some of
the electric vehicles actually fast charge is not so damaging as we thought it's three to five years
ago so my my advice is to to use your cars as you would like and yeah I think that should fit
your purposes rather than you you worry about maintenance of the car I think that's a lot of
the design thinking the the car manufacturers put into as well and at at the moment fast charge is
is not like if you are doing 20 minutes charging it's it's actually not too damaging to your car
unless you kind of do it 10 times per day which I don't think I would want to do that yeah okay I
guess and I guess because most fast charging as well it's all being managed by this stuff and most
of it is going to be from a low number to a 75 percent or something in which case it's all quite
it's all managed it's all within sort of spec and one thing that I have heard is bad
is leaving a car fully charged in the heat is that particularly bad
um so that will in in theory if you have a laboratory test that will cause
something we call calendar aging which means you lose some some level of capacities
if if you were to leave a battery fully charged and leave it out there but I I
I think the truth is actually inside the battery management system this is being managed as well
um so while you are so it's it's trying to calculate um the level of sort of how full the
battery is when you kind of leave it too long it will try to navigate to so maybe keep it a bit
lower or keep it a little bit more lower depending on the the duration of storage I I'm just giving a
very simple example but in reality that there are algorithms inside the battery management system
that kind of take care of this so yeah I wouldn't say that's too worrying
and is this the case this this level of control and management has this been around
for 10 years or is this just kind of what's in if you buy a car new now
so this has been around in the cell phone and laptop industry for more than I would say if you
have a laptop you pretty want to press a health mode button and if you irrigate your battery you
can sort of have health mode and if you let the sort of device know about it they will try to
change the management algorithm so it anchor towards sort of healthy and better durability
so that's the learning that people gather from a consumer's electronic device
market and transfer to the EV market I believe okay yeah yeah yeah so people I mean I know with
my phone it's like oh we're only going to charge you know you can now pick oh I've got an iPhone
and it's like you said it so I don't know I think I charged at 85 or 90 and I never really charged
to 100 but yeah companies have been doing this for a long time and as you say like
an EV manufacturer they don't want their battery to die in five years so they're going to try and
hedge it a little bit um because if you have a battery pack you don't get there's a there's a
usable number and the full number why is that is that because of this management stuff or is that
just when you create the certain size you just can't fill it to that much it's mostly because
this management strategy basically the when you are going from a almost full to the
fullest states the management in there needs to be more conservative so in in in what we touch on
like closing the the tap yeah so you need to close the tap quite carefully so you allow the user to
navigate in that region but you don't make it overflow in innocence so that part in in a sense
needs more careful management and the user benefit of navigating that may not be so
like beneficial if that becomes your day-to-day usage yeah so that's I think the design thinking
going behind a usable mode and a full mode okay cool well thanks very much for coming on the
podcast and giving me a bit of a dive into batteries and where it's at with all of the
technologies yeah thank you very much for inviting me as well cheers
this podcast is part of the sports social podcast network
you
About this episode
Jingyi Chen from Breeze Battery Technologies discusses the complexities of battery management software and its role in enhancing electric vehicle (EV) performance. The conversation dives into the intricacies of lithium-ion batteries, the importance of real-time data for optimizing charging speeds, and how software can mitigate issues like lithium plating. Chen highlights the evolution of charging technology, the significance of offline simulations in battery design, and the collaborative future between battery manufacturers and car makers. Insights into fast charging practices and battery longevity are also shared, making this episode a deep dive into the future of EV technology.
Jingyi Chen, Head of Battery at Breathe Battery Technologies, joins me to discuss how software is transforming the way EV batteries are charged, managed, and designed. We get into the science behind fast charging, battery degradation, and why Breathe’s tech is already being used by major OEMs like Volvo, Rimac and Cosworth.
Enjoy.
https://www.breathebatteries.com
Chapters:
00:00 - Intro: Smarter batteries, faster charging
02:48 - What makes lithium-ion batteries tick
05:44 - Battery management software: the brain of the pack
08:30 - Real-time monitoring, diagnostics & control systems
11:36 - Lithium plating, fast charging & what can go wrong
14:38 - Where charging tech is headed next
17:31 - What's coming in future EV battery tech
20:23 - How to treat your battery to make it last
22:27 - Wrapping up: challenges and breakthroughs ahead
Learn more about your ad choices. Visit podcastchoices.com/adchoices