Modular Electric Powertrains

Today we are joined by Mohamed Badawy, Founder & CEO of Scalvy Inc. He and Zach have a fascinating conversation about Modamed's company, Scalvy, and its innovative powertrain technology. As a part of this chat, the two cover how long it takes to introduce new concepts to the market, how Scalvy's tech works, AI integration, and much more. 

Listen to the Episode:

Download this episode.

Watch the Episode:

Show Highlights:

• Intro
• Mohamed & Scalvy's Background
• Mohamed's Research
• It Takes Time to Explain a Concept
• Modularization & Competitive Advantage
• Safety, Reliability, & Auto Makers
• Automotive Powertrain Standardization?
• Creating the Powertrain Standard 
• Maintenance Costs?
• What Happens with the Powertrain
• AI Integration?
• Aerospace Industry
• Marine and Agricultural Vehicles?

 

Transcript:

Zach Peterson:

I'm sure as soon as you start bringing up a software in the vehicle, someone is going to throw AI at you and ask you how you're integrating AI into this. Of course, it's a hot topic. It's been the hot topic for the last year. So I have to ask, is there AI involved or will there be AI involved in what guys are doing?

Mohamed Badawy:

So AI was involved in control algorithms and control structures for as far as I remember. It just never had this hype. So yes, every single control algorithm we use, it probably has an AI component into it, but it wasn't named AI and it didn't have this hype around it. So for example, you find that we're using a lot of model predictive control. They call it model predictive control because it's predicting what's going to happen and then it's taking a decision based on that.

Zach Peterson:

Hello, everyone and welcome to The Altium OnTrack Podcast. I'm your host, Zach Peterson. Today, we're talking with Mohamed Badawy, founder and CEO of Scalvy. He's also an associate professor on leave of electrical engineering at San Jose State University. As I said earlier, he's the founder and CEO of Scalvy, a startup that has just been brought into Altium's launchpad program, and we're very happy to have him here today to discuss his company. Mohamed, thanks so much for being here with us.

Mohamed Badawy:

Thanks for having me, Zach. And great to meet you.

Zach Peterson:

Yeah, great to meet you too. So I think the first thing to do is maybe tell listeners a bit about yourself and then a bit about Scalvy.

Mohamed Badawy:

Yeah, sure. So Scalvy is a startup that was born out of my research. The idea of Scalvy is how to modularize the electric mobility space. So we have been through a phase where electrification of mobility fleets was a niche, but right now we're at that new stage where that's really the trend of the whole market. We're trying to electrify everything from passenger vehicles to electric trucks to electric buses to marine applications to army applications and so on. So the way that we have been doing things have been building customizable, centralized systems for this electrification space. And right now what we need to do is how to modularize these systems so that we can develop faster at a lower cost, at a faster rate, and at a better performance. And that's what Scalvy is trying to do.

Zach Peterson:

So you said this originates from your research. Maybe tell us a bit more about your research.

Mohamed Badawy:

Yeah, sure. So I graduated with my PhD in 2016 from the University of Akron in Ohio. And as soon as I graduated, I remember I graduated on a Friday. On Monday I was working as a professor at San Jose State University, which is the biggest public university in Silicon Valley. When I arrived at the university right away, I founded and directed a research lab focusing on electric mobility applications. My focus there was building modular converters, how to bring modularity to power conversion systems, and particularly in the electric mobility space.

So that was my main research for a number of years. What happened is that one day I received an email from a large army contractor asking me if I know of any company working on modular powertrains, modular electric powertrains. And when I asked why, the answer was, "Well, in the army and in the DOD in general, we have so many vehicles that we need to electrify." But every one of them has its unique features, has its unique requirements.

And for us to build customizable solutions for every one of them is a daunting task. So if there is a way that there is some sort... And he didn't really know how to articulate it, but he was thinking if there is a way to build small modular converters or something and put them together to make a powertrain, that would be ideal. And that struck me because that's what I'm working on. I just didn't think about it for this application. And then I started seeing how modular converters and modular electronics and distributed control can be applied in the electric mobility space.

I started first as a research projects and then it grew out of the lab, and then I needed to really leave the university, leave my research lab and start working on it as a separate entity and as a startup. That's when I joined forces with my co-founder, Amr Ibrahem, who was at the time working at Delphi, BorgWarner, which is the largest powertrain company in the US for the past seven years or so. And then we joined forces together and we started working on Scalvy.

Zach Peterson:

So the military is interesting, and it's interesting that your idea came from a question from someone in the military because they already use some level of modularization when it comes to the chassis. What took them so long to, I guess, figure out that they might need to do this with powertrains?

Mohamed Badawy:

Yeah. Certainly there is... Because when you talk about powertrains, let me try to define what we are actually working on. So in any electric mobility application or electric vehicles, we're particularly talking about battery powered electric vehicles. So you have the battery and they have all the electronics that's going on, the battery management system, the onboard charger, the inverter or the motor drive system, the DCDC converter, all of these are working together controlled to be able to drive a motor and regulate a DC output and charge the batteries and protect them and all of this.

So all of these electronics to modularize. It's quite a challenging task. It includes a lot of research and innovation to be able to do that in an efficient manner and most importantly in a cost-efficient way. Because in the mobility space, people are trying to reduce their cost, not increase the cost of their components. So it's a heavy research topic.

So it's not surprising that it doesn't come immediately to someone that, "Oh, I need to modularize this." When people are used to a certain way of doing things for so long, Tesla have been doing it like this. All the companies that electrified. The first EV from GM, it was a central unit. People think that's a way to go. And then they start building everything and the standards and everything around the way that they use to see how it's progressing. So when we came into the scene and saying, "Okay, there is another way of doing it. It's actually safer and actually more reliable," and so on, even when you look at it, when you just look at the configuration, it looks quite simple, but it takes people some time to register how this is actually happening.

How can you completely get rid of the DC bus that was the heart of every single electric vehicle? What do you mean you don't need it? How come you don't have a large inverter? Who's going to process this power? How come your powertrain modules is acting as your BMS modules? That's strange. But then as we explain more and as we go through the technicalities of it, it becomes clearer and clearer to them. So it's not surprising I would say that it takes some time to understand the concept or to register the concept.

Zach Peterson:

So how have potential customers or partners responded to the modular power distribution concept? Have you gotten positive responses?

Mohamed Badawy:

The same way that I was explaining now, yeah. So it's actually the same way that at the beginning, "That's surprising. That sounds weird. But oh okay, so that's what you're... Oh, okay." So that makes sense and so on. So the biggest thing is that it depends on the type of customer. So for example, low volume OEMs, what are they suffering from? They're suffering from the lack of interest from suppliers to work with them. So if you're developing 500 vehicles or a thousand vehicles or a couple thousand vehicles a year, the large suppliers, they're not interested in you.

The large suppliers, they're only interested in the large numbers, otherwise it's not going to make economical sense for them. So solving this problem for them would be the main target. Then there are another set of customers that are working, all they care about is the size and the weight of the powertrain. If you're able to decrease that, if you're able to decrease the weight for me, whether that's for aviation vehicles or we're working with UAV customers, unmanned aerial vehicles, that's the most important thing. If you're able to reduce that, that's the biggest value proposition to them.

Others, it's all about safety and how can you increase the safety and the reliability of my system. And fortunately with modularity comes redundancy. With redundancy comes safety and reliability. So we're able to do that for them. So what I would say is that whenever you take any concept and you try to modularize it, you try to build a platform around it. A lot of advantages are going to start presenting themselves. Advantages that when we started this project or this company, we were not counting for them. We didn't think that we're going to talk to people, telling them, "You know what? You can extend your range if you use our powertrain."

But as it came about, as it appears, you actually can. So what I would say is that as every industry matures, the need for modularization arises as it modular arises, a lot of other advantages or benefits are unlocked. And that's what we're seeing in our conversations and our early commercial traction with our customers.

Zach Peterson:

So I guess modularization, excuse me, could happen in two ways, right? There could be vendor specific modularization where one automaker, let's say they modularize their entire powertrain so they can mix and match components in different vehicles, but then there could be something that's standardized across multiple automakers.

Mohamed Badawy:

True. Yeah, very true.

Zach Peterson:

Which of those do you see being the path forward for car makers to ensure that they can stay competitive, really jump into the EV space and provide these new vehicles for people who want them, but then also ensure that they maintain competitiveness?

Mohamed Badawy:

Yes, yes, yes, for sure. It's a mix of both. It depends on the size and the goal and the perspective of the customer. Usually large volume OEMs, the largest companies, they want to have this competitive advantage. They take a lot of pride in building our own powertrains meaning that we are the ones who are designing the main architecture for it. For these kind of customers, yes, certainly building modular solutions that can be used across different models of their vehicles would be a huge advantage for them.

Rather than you waiting two or three years to electrify every single model, I'm able right now to use the same technology across different models and different platforms because we call our technology, an under the hood technology. It's not really affecting the drivability of the person driving the vehicle except for the range extension and for the faster response. But it is quite technical and quite critical in building your powertrain.

Now, the other set of customers are customers whom my goal is to build an autonomous vehicle. My goal is to build a UTV for these certain applications. My value proposition is a chassis of my vehicle and how durable it is, or it's about the autonomous software that I'm putting into it. So they don't really care about having this competitive advantage in my electronics level. And these tend to be usually low volume OEMs.

With this set of customers, what we're trying to do is to build the same hardware, the same platform board, which I'm showing right here. That's kind of a version of what we're building. Building the same platform and the same hardware board across all of these customers and work on the most restrictive safety standards among all of them so that we can reduce your cost. We are able to manufacture a high volume even though every one of my customers is a low volume OEM.

Zach Peterson:

So you bring up safety and of course the related topic reliability. I'm sure trying to work with the big automakers creates a lot of challenges for proving reliability. They're a pretty conservative industry and they want to make sure that when they put out 10 million cars have your equipment that all 10 million cars are going to be reliable. So whose job is it to prove that the modular approach really is reliable and what does that runway look like?

Mohamed Badawy:

Yeah. So that's a great question. So that's why in our strategy, we're not planning to bring in one of these large auto makers to be a customer of ours in the first couple of years. That's not the plan. That's not our path moving forward. Our path is to work through their preferred channels. So whenever you talk with any large automaker, they have a set of tier one suppliers that they have been working closely for quite some time. The tier one supplier expects what's going to be the requirements of the next vehicle of this OEM and the OEM knows the capabilities of this tier one supplier and they trust what they can provide to them. We are getting into this channel. We are getting our configurations and our technology and our software algorithms to be part of this channel so that we are able to provide our systems through their preferred tier one suppliers.

So it then becomes our job, the tier one supplier job and the OEM job and the standards and the way that OEMs are working with suppliers and with software providers is defining this process well for us. But it's an ultimate job for all of us together. So that's with these large automakers. With the smaller automakers, we can take a little bit more of a leading role in this relationship, given that most of these automakers are looking for a system solution that can easily plug and play into their vehicles, and that's what we're providing to them.

Zach Peterson:

So when you say a smaller automaker, that might be a newer automaker, maybe Rivian for example.

Mohamed Badawy:

So it could be a newer automaker or it could be a large automaker, but is working in an inherently low volume space.

Zach Peterson:

Okay.

Mohamed Badawy:

So if you look for example for off-road vehicles or UTVs or all of these kind of applications even though you could be like a multi-billion dollar company, but every model of those, you're only going to produce maybe thousands of it every year. It's not going to be in tens of thousands of units.

Zach Peterson:

I see, I see. So in terms of the reliability challenges just at the electronics level, what are the things that you have to consider in some of those designs? I know with auto it's high vibration, variable temperature. What other approaches and what's your strategy?

Mohamed Badawy:

Yeah, certainly you do a lot of stress testing, vibration testing as you mentioned. But I would say overall that you are looking for two things. You're looking for... Okay, I have these small modular boards and I'm connecting every one of them to one of my battery modules. That's basically our configuration, our main architecture. It looks something like this. Every one of those is connected to one of my battery modules. So now you have two tiny-

Zach Peterson:

Sorry, sorry, if I could just jump in.

Mohamed Badawy:

Yes.

Zach Peterson:

So you have a PCB inside of a small enclosure. It almost looked like it was 3D printed.

Mohamed Badawy:

Yes. So that's actually a representation of how the final enclosure would look like. It's a representative way of how our enclosure should look like that we're working with our partners on. And that's a PCB. That's only one of our designs. We have around 10 different modules that we're offering. So then it becomes... The question is, "Okay, what kind of fault that can happen?" So it can be either something in our modules, this can happen, or it can be in the battery module that we are communicating with and that is supplying powertrain module which we call SIM, smart integrated module.

And then it becomes, "Okay. What's my strategy for every single fault?" If one of my battery modules is faulty or it's discharged or it's overcharged or anything that happens to it, how can I protect it? So you have to have all of this in your hardware configurations and in your software algorithm. Also, every single MOSFET here, every single processor here, if any of this is going through a fault, what's my strategy to overcome that? So that's where you start working with the safety standards such as ISO 26262 or R100. And that's where you start defining what's going to be SLD, SLC, SLB, SLA, and these are the different kind of SL level, safety levels that you are assigning to every single subsystem in your solution.

Zach Peterson:

And then with modular components, they have to hook together in some way that they have a common interface that they can all communicate with. So this goes back to the vendor specific standardization versus cross vendor standardization. I guess this next question is, has the automotive industry moved at all to try and standardize power interfaces across multiple components or equipment? It sounds like they haven't.

Mohamed Badawy:

They haven't much. They're trying. So almost in every single electric powertrain company. You're going to find marketing material on modular powertrains, but when you dig a little bit deeper into every one of them, you're going to find that what they mean by modular powertrains is that we're going to give you powertrains at five or six different sizes, and then you're going to choose one of them. The problem with this strategy is that these suppliers are then forcing me to choose a size and then design backwards.

So the powertrain, which is maybe 10 or 15% of the cost of my batteries, will determine the capacity of my batteries and the power of my batteries. And that doesn't make sense. I should first determine what's my capacity, what I want my vehicle to be, and then find the powertrain that fits exactly into this. That's going to give me the voltage I want, the current that I want, the DC output that I want, the onboard charging levels that they need and so on. And that's not currently available in the industry except through the highly customizable channels, which we talked about earlier.

But if you look at every single one of these suppliers, you're going to find that they are having what they call a platform team or a platform engineering team that is trying to figure out how to use our legacy products to build a modular platform. And this is the main issue is that they have to use their legacy products and their legacy designs, which are inherently customizable, and we are offering an outlet of this.

Zach Peterson:

They almost have to throw out all the old stuff and totally approach it differently in order to get any kind of modularization. It just sounds like the way that these companies are structured, they're not prepared to do that.

Mohamed Badawy:

Yeah. And that's why even right now in our conversations with suppliers, because we're seeing ourself as a company that can be complimentary to the work of these suppliers as our technology can support them. That's why in our conversations with them, they need us. They need us to drive this forward. It's quite costly for them and it would take years honestly, to be able to figure out how to do all of this efficiently, the control, the configurations, and then build your IP.

But if you have a company that has that provided already, then it would make a ton of sense for these companies to partner with us. And that's why currently as we speak, we have these conversations ongoing with a number of the largest suppliers in the automotive space.

Zach Peterson:

It sounds like this is a good opportunity for a company like yours to really create the defacto standard for power interfaces throughout the powertrain, and then that really creates a motor around your company.

Mohamed Badawy:

And that's our main goal. So our main goal is how to get to the market as fast as possible and as efficient as possible using the high safety protocols. And that's what we're doing right now. That's why our team is working extremely hard to be able to get our footing on the ground and test our system with the early samples that we started working on with our customers, and then that's going to get us into production.

Zach Peterson:

So switching gears on this just a little bit, no pun intended, how would your type of system or a modular approach to powertrains affect the maintenance costs for vehicles?

Mohamed Badawy:

So we have worked with the Air Force in the past through one of the grants that we got, and that was the biggest selling point, is that they were telling us, "Okay. We want to electrify, but no one is ready for it. No one is ready for how to maintain an electric powertrain. If the inverter blows up, "What are we going to do? Is it going to be a completely new project that we have to replace it with another one, go to the supplier? If there's something that we need to debug in the code, something in the code that is not working, what are we going to do? If my battery management system has some issues, how am I going to handle it?"

And then the solution that we're providing to them is that you can treat these powertrain modules basically as fuse boxes that you put in and out. So if you have one of your powertrain modules need maintenance, you just take it out, replace it with another one, and that's it. So that really makes it that much easier to maintain your powertrain. It's not just about the batteries, it's also about all the electronics and the software that goes in. Additionally, all of our software algorithms can be accessed over the air and can be periodically updated.

Even when we come up with a more efficient algorithm or with a better control structure, we are able to upload it to our customer vehicles for the period of about 10 years after supplying our powertrains to them.

Zach Peterson:

So you're talking about over the air updates for embedded firmware and software. What exactly does your system need to do within the powertrain, just in terms of what happens at the embedded level? You brought up control a couple of times, so I'm a little curious about that.

Mohamed Badawy:

Yes. So on the firmware level, I like to think about our modules as three different layers. The bottom layer is the battery management system. How can I protect my batteries? How can I balance them? How can I control their charging and discharging profiles? All that is happening with the battery management system and all the sensing signals. And then the layer above that is the power electronics layer. How can I get the power of the battery converted into AC so that I can drive the motor, convert it into DC for my auxiliary loads? How can I accept charge to charge my batteries and so on.

And then the last layer is the control and the software algorithms. And this is the layer that's controlling all of these operations. So this is basically the supervisory layer above all of them that is telling the Beam S system what to do. It's telling my own onboard charging what to do. It's telling my inverter and the DCDC converters what to do. We're doing all of that locally, meaning that every single board of those is equipped with its own controller, its own processor that can take decisions locally based on my local measurements.

And that gives us a huge advantage in terms of our response, our response to any faults that's happening. I'm not waiting for the large controller, the supervisory, the master controller to tell me what to do. I am able to take decisions here locally. And also there are some other supervisory controllers that are controlling more of a higher level control algorithms over these modules, and that's it. So it's basically the software algorithms are the vehicle that's running everything happening on this module.

Zach Peterson:

I'm sure as soon as you start bringing up a software in the vehicle, someone is going to throw AI at you and ask you how you're integrating AI into this. Of course, it's the hot topic. It's been the hot topic for the last year. So I have to ask, is there AI involved or will there be AI involved in what you guys are doing?

Mohamed Badawy:

So AI was involved in control algorithms and control structures for as far as I remember. It just has never had this hype. So yes, everything control algorithm we use, it probably has an AI component into it, but it wasn't named AI and it didn't have this hype around it. So for example, you find that we're using a lot of model predictive control. They call it model predictive control because it's predicting what's going to happen and then it's taking a decision based on that. So that's an AI tool. It just didn't have the hype of an AI tool. Also, there is a ton of applications in the battery management system for AI.

How can I model my battery management system? The whole battery management system is revolving around AI really because you're building a model and you're making a ton of assumptions based on your online calculations and online measurements of your battery cells and battery modules and all of that is done autonomously. So there is a ton of AI involved here. Also for maintenance, for predictive maintenance to be able to predict if this battery module needs to be changed or if this battery module is faulty and they need to do something about it, our A algorithms are able to predict that.

Finally, for the second life batteries because there is one thing that we need to talk about is what's going to happen to all of these batteries after they retire. They still usually have 65 to 75% of their capacity or maybe up to 80 and in some cases even more. So what are we going to do with them? And our algorithms and our powertrain modules are able to identify exactly how much capacity you have in each, what's the state of health of each one. Not battery pack, but battery cells and battery modules so that when it goes to its second application, the company is working on it have the full data about this battery modules through our battery management system algorithms. And all of that is done also enabled using AI.

Zach Peterson:

That's interesting. I had not heard of second life batteries, and it sounds like maybe once it's done in the... Or the batteries have been used up to their capacity limit. In the first vehicle, they maybe get resold to some other company maybe that's doing a smaller vehicle, let's say. They don't need a brand new huge battery pack.

Mohamed Badawy:

Yeah. The most primary application for them had been in using them as energy storage units with solar or with wind, so that they can support the renewable energy penetration into the grid. Because the more renewable energy we're bringing, the more storage we're going to need because of the disturbances that can happen with solar panels, wind energy.

Zach Peterson:

Okay. I think that makes sense. So you've mentioned a couple of industries so far. I mean, obviously automotive, consumer, commercial, and then you even mentioned stuff like off-Road, but you briefly mentioned aerospace. What are some other industries where this type of modular approach to power management and distribution could be taken?

Mohamed Badawy:

Yeah. So a few ones that come to mind that we had success in attracting customers in them. One of them is aviation. eVTOL and UAVs. eVTOL is electric vertical takeoff and landing and unmanned aerial vehicles, all of this, we have applications right now that are ready. So eVTOLs, when we think about it, we think about companies such as Joby Aviation or Archer or others that are working on passenger vehicle eVTOLs, and we think it's something of the future.

But right now there are applications for eVTOLs in farming, in monitoring, and then other spaces. So these kind of applications, our approach is attractive to them for two reasons. We're reducing the weight of the powertrain and anything that flies weight comes as the number one design factor. And the second reason is that all of them are still producing a low volume. And really in low volume our system prices are not even comparable to either off the shelf or to using customizable solutions.

So that's one application. Another application is off-road vehicles. That's used for construction or for mining, or for farming. We really have to think about this sector and how can we electrify it. They are producing enormous amount of greenhouse gas emissions and we really need to electrify this sector.

There are some challenges there. One of them is that in off-road vehicles you might need like three megawatt, five megawatt or even more in a single vehicle because they're doing heavy duty work. How can you achieve that using the current solutions? You can't. You just can't. Unless you're going to be using a modular approach such as ours, and that's why we're having success in gaining traction and attracting customers in this space.

Another one is marine applications, which is kind of falling into the same need as off-road, looking for heavy duty, looking for long haul operations, looking for high power applications that the current semiconductor devices simply cannot provide. One of the main factors that I expect to drive our technology forward in the future is when people realize that, "Oh, to use efficient semiconductors at high power, high current, high voltage ratings, really we need to use modularity because the other best option in the market right now is the use of silicone carbide MOSFETs. And silicone carbide MOSFETs are great, but they pollute a lot in their production process.

They need to be heated to very high temperatures. They produce a ton of CO2 emissions for every kilogram or every ton produced of silicon carbide. And that's something that's only going to be evident as scale of electrification goes up.

Zach Peterson:

Okay. That's very interesting. I would imagine that getting into maybe the marine space or getting into agricultural vehicles, that's going to be maybe a little bit of a lower barrier to entry, both in terms of engaging with suppliers, but also in terms of proving reliability. Would you agree?

Mohamed Badawy:

Yes. A hundred percent. And that's why that's our primary target, and that's why hopefully when we make a public announcement about our production run and about our first set of customers, most of them are coming from these spaces.

Zach Peterson:

Well, as all of that unfolds, we would love to have you back to talk about it because this sounds extremely interesting and I have to be honest, I'm a bit surprised no one thought about it earlier.

Mohamed Badawy:

Yeah, we got that a lot. That's a lot. "It looks simple. Why no one did that in the past? Well, okay, we're telling you. Here you go. You can be the first one to do it." We tell that our customers.

Zach Peterson:

There you go. Great. Thank you so much for being with us today.

Mohamed Badawy:

Thanks so much, Zach. I appreciate your time today.

Zach Peterson:

Absolutely. To everyone that's out there listening or watching on YouTube, we've been talking with Mohamed Badawy, founder and CEO of Scalvy. He's also an associate professor on leave of electrical engineering at San Jose State University. If you are watching on YouTube, make sure to hit the subscribe button, hit the like button. You'll be able to keep up with all of our tutorials and podcast episodes as they come out. And last but not least, don't stop learning, stay on track, and we'll see you next time. Thanks, everybody.