Texas Instruments’ General Manager of Grid Infrastructure and Power Supply, Henrik Mannesson, joins us on the Moore’s Lobby podcast to discuss the challenges and opportunities associated with real-time energy measurement and management in the age of renewables.
Optimizing energy generation and consumption requires accurately measuring currents and voltages. In addition, to maximize overall efficiency, that data must be shared in real-time or near real-time.
Texas Instruments aims to design a smarter, more efficient grid and renewable energy system. Image used courtesy of Adobe
The highlights of this conversation between Mannesson and host Daniel Bogdanoff include discussing:
- The differences between power management in the home or small factory and power management at the grid.
- The evolution from smart metering to energy management.
- The importance of accuracy in power measurement and how to achieve it.
- The benefits of staying with a single company for many years.
A Big Thank You to Our Sponsors!
Meet Henrik Mannesson
Henrik Mannesson has spent nearly two decades with Texas Instruments, where he is currently the General Manager for Grid Infrastructure. His team helps customers solve design challenges in smart meters, solar energy, EV charging, and grid automation. They aim to make the electrical grid more efficient, more intelligent, and more resilient.
Mannesson holds a Master of Science in electrical engineering and a Bachelor of Science in business administration from Lund University, Sweden.
The following transcript has been edited for clarity.
All About Circuits’ Daniel Bogdanoff: From EETech Media and All About Circuits, this is Moore's Lobby. I'm your host, Daniel Bogdanoff. Today in the Lobby I’m joined by Henrik Mannesson, General Manager of Grid Infrastructure and Power Delivery at Texas Instruments. He works alongside his team to help customers solve design challenges in smart meters, solar energy, EV charging, and grid automation with the goal of making the grid greener, smarter, and more resilient.
This episode is adapted from one of our Industry Tech Days keynote interviews, but if you want to learn more about the world of power, you are in for a treat.
Entry Into Technology
DB: Henrik, thanks for being here today. I’m curious to hear how did you first get into and excited about tech?
Henrik Mannesson: So I think it started fairly early. My father worked on air navigation systems back in Sweden. He worked for over 40 years in that field. That also meant that we had computers and electronics at home from a very young age. So I can remember being in the basement and trying to put things together (or maybe more damaged things!) when I was younger.
Then, somewhere in the mid-80s, computers started to come into it, and we spent a lot of time programming computers and things like that. So that's probably how I got into it.
I think what excites me about it is that you can take a limited problem and solve it in a small piece, and then you add multiple of those things together and you have actually solved a really big problem. So I think that's maybe the really exciting part of electronics and what we do here.
DB: Do you remember the first computer you had? And was there a time you just totally broke something tech at home, that you remember?
Mannesson: I remember spending a lot of time building audio codecs and so on for those computers that you could record audio and get audio out of them. That was something I remember. This was at a time when I think audio was a media files type of thing. Iit was not really a Spotify quality of audio that came out of those computers. But I remember back then, we broke a lot of really expensive A-to-D and D-to-A converters trying to get this to work, but it was a really good time.
DB: So you had to come to Texas Instruments to get an unlimited supply of A-to-Ds and D-to-As?
Mannesson: Yea, maybe!, I think actually it was probably Burr-Brown audio codecs. They were probably the top-of-the-line codecs that you could get at that time. So I'm sure they were actually or it later became Texas Instruments at least.
DB:So then as you got older, you went and got an undergraduate and master's degree in electrical engineering and also an MBA. Can you tell me a little bit about what it was like doing engineering and then also adding in the business side of things?
Mannesson: I think it came down to that I felt like to just solve technical problems, maybe it's not always enough. I think if we are not able to make it economically attractive as well, and I think if I look at that today, how that plays out today in the field that I'm in today, this is in solar generation has been around for at least 15 years. If we now look at it over the last two, three, fourm or five years, it's probably at a time point where it has become really economically viable as well. That is actually also what accelerates the adoption today at a much, much larger and much, much faster than any technology advancements alone could do. So I think that combination is really interesting.
Early Career and Manufacturing
DB: And you've spent all of your career, it seems, at Texas Instruments. How did you first find TI and get started there?
Mannesson: So I started off with a training program at TI. It was called a European Graduate Program but it was basically a training program where you got to see different parts of TI. So I was hired by TI in Sweden. I went then to the UK for six months and was in Germany as well for another six months for TI, working and seeing different areas of the company.
DB: Is there anything that surprised you going from grad school into industry?
Mannesson: I think maybe it's how to get things done, I think in college you always think, or you tend to thinkm that the most important thing is to get your circuit right, calculation rights, and so on. And that's how things move. I think when you start in a large corporation you realize that how to move projects forward at the customer, understanding the difference that your company can make, and how our solutions fit into that, that's a much larger question and something that's much harder to answer than to get your calculations and circuit right.
DB: That's where I guess that business thinking matches up with the engineering thinking; that's, I think, a very critical combo skill set to have.
Can you tell me about your move from TI, Sweden to TI, Germany? I understand there was a change in responsibilities.
Mannesson: I moved to Germany to get into a more pan-European role. There was more exposure in that role as well. It was at a pan-European level, of course. So you got to see more different types of customers as well.
I was supporting the industrial market for TI then, and especially then being in Germany in the heartland of factory automation and industrial automation of Europe was really exciting. So that's what triggered that move.
DB: And you worked as a system engineer and manager and for factory automation systems. Can you tell me a little bit about that space and what you did there?
Mannesson: That was much more of an engineering role. We were trying to help customers in that factory automation motor drive space, At that time, I think this was a time where industrial Ethernet protocol and so on really started to take hold in this market. So that was probably some of the most exciting things we were working on then.
You can see that today—how basically all data in a modern factory today gets digitalized very early and then used to control very complex processes. It was also a time I think where functional safety, collaborative robots, and things like that started to emerge
Some of those things, like functional safety today, are almost table stakes for some of those markets. But back then, when I was there it was very early into it, so it was an exciting time.
DB: I think this was the 2014 to 2016 time frame. What was state-of-the-art then, and what has been interesting to watch play out to today? Because I understand smart high-tech manufacturing has had a lot of advances in the last ten-ish years.
Mannesson: I think it was really where it's been. Let me give two examples there. Maybe one on the Ethernet and connectivity side where you start to digitalize the data right out at the sensor or very early, far out in edge nodes of the system.
And then be able to spread that data out and use it very effectively to control a process flow in a factory. So that was then very new and new concepts. Today I think if you look at it, this is really where any modern factory actually is going today. That's of course exciting to see how that has played out.
DB: Are there any projects or automation challenges that you're especially proud of solving during that time?
Mannesson: Difficult question. I said I think what we worked on there on functional safety was actually exciting because this was at a time where functional safety requirements started to emerge in this market. Functional safety is really, really a systems challenge.
So I think our team then was in a very good spot to spearhead that and start to solve some of those challenges. Because functional safety, you can solve it for hardware, you can solve it through multiple different types of hardware, you can add software to the mix as well, you can do two different channels and compare data, and so on. So there are many different ways you can achieve functional safety level and those are some of the system challenges that I worked on solving back then.
And I think there are only more of those types of challenges today. If you think about things like collaborative robots and things like that, where we actually are trying to put machines very close to humans. So of course you need to be very careful when humans and machines are going to collaborate on the same factory floor.
DB: Yeah, I always understood that in concept, but last week I came across one of those Boston Dynamic Spot robot dogs in person. It was walking down a hallway towards me, and I felt like legitimate terror. And this is just this little robot dog, not a giant manufacturing arm or something like that.
Mannesson: I don't think you are really maybe putting giant manufacturing arms directly to humans right now. But I mean, you see those more collaborative robots that are maybe helping out with some parts of assembly at assembly lines and so on or do mechanical things or when you need to lift heavy objects and things like that. So that's definitely where those things can help and assist.
Of course, that safety aspect of that has become very, very different compared to having a large manufacturing robot that you can gate in somewhere, and you don't let any humans come close to it.
Transition to Grid and Power Delivery
DB: Another area where I see safety popping up in a lot of conversations is with grid and power. In 2016 you moved from Germany to Dallas, Texas to become GM of the Grid and Power Delivery Systems group. Can you talk me through a little bit of why you made that change, and then also as a former Texas resident, what was it like moving from Germany to Texas?
Mannesson: So I know, from being in that role in Germany and the systems manager role that I was in there, I saw the rise that came on for grid infrastructure systems, especially solar, EV charging, energy storage systems, those types of things, and how much of impact that's going to have on the grid networks over the coming years. So that was very exciting to me. That's one of the reasons then that I accepted this role in Dallas and joined TI over at its headquarters.
I think if you talk about the difference, TI is I think a fairly similar company across the globe. They act in a fairly similar way. If you are in Germany or Sweden or in the US, I don't think it's that large of a difference. So I would probably describe it as that there is a larger difference for the family maybe than for me moving. Because I'm still spending a lot of time inside TI. We actually work a lot today with my previous colleagues out in Europe as well. So I get to spend a lot of time with them even from here. For my family, it's probably a larger difference than it is for me.
DB: What are some of the areas that TI is involved with and how are they part of the design process for smart grid and power?
Mannesson: So I think if you start there, there is an energy transition going on in electrification. I believe that there are four large areas inside there where semiconductors are going to create a large difference over the coming years.
One of those areas is in high-voltage power conversion. So that's any AC-to-DC, DC-to-AC conversion, and that centers around real-time controllers. It centers around things like wide bandgap devices and so on. But to control and convert power between AC and DC.
Then the other thing is current and voltage sensing where also the requirements are increasing, both because we need to measure the power in many more places, but it also needs to make the systems safer to operate. So that's also an area where current and voltage sensing comes in.
The other thing is then edge processing and the connectivity. So by adding both edge processing and connectivity options to make the equipment thereby more intelligent, that's going to help a lot on how to decide what to do with this power that we now have generated or want to consume.
The last thing is really the battery management side of things where I think monitors, balancers, and so on, has a huge impact on how we're going to be able to deploy batteries across the grid network, but also in people's homes.
DB: I was taking a walk down our street the other day and I looked up at the light pole and I noticed this box with antennas coming out of it, and I got into this whole deep dive of like, what does the metering system look like and the infrastructure for that. I was fascinated how high-tech power metering is and how high-tech smart power metering is. Can you tell me a little bit about what that ecosystem looks like and your involvement in it?
Mannesson: I think smart meters started to deploy maybe around 2008, 9, or 10 or something like that. That's the first generation of smart meters. I think what semiconductors were able to do then was to move a system from mostly a mechanical E-meter before that to now become a semiconductor-based E-meter.
Those E-meters and what we got by that was a lot of it was radio communication. So we're able to record data in the meter and then upload it to a utility backbone network every so often so that we could do accurate billing, basically what we call AMR, so automatic meter reading. So I think that was where we and that was probably where we were back in 2016 when I joined this group.
DB: Yeah, I remember meter readers being a thing, but I sure haven't seen one in a while. It's clearly also getting more complex. There are solar panels and cars that can potentially do vehicle-to-grid. There are a lot of factors playing in here. As we look ahead at the expansion in renewables technology, how are the metering and control requirements changing?
Mannesson: So I think first and foremost it comes down to net metering. As you said, you're going to be able to sell, I mean, you can sell back energy to the grid now, so you need to be able to measure the energy in both directions. Both what you draw from the grid but also what you provide back to the grid.
Then of course as well, you probably in the future would want to have an E-meter and a system in your house that has knowledge about the price on the energy market and do that much more in real-time than it is today. I mean, today the meter actually doesn't know anything about the price of the energy at any given point.
Over time here, as we go through, actually what's going to happen is that prices are going to start to become more dynamic because we're going to have to line up better consumption and generation. When we do that, we're going to have to measure and upload data much more often. So I think those are things that we from, Texas Instruments, where we spend time right now so that could be accurately measure that energy.
We have analog front-end solutions and microcontroller backends that then can run software to make those energy calculations and then connect to radios that can upload this data much faster and much more often than you do it today.
DB: Technologically the infrastructure is a lot more than just measuring voltage and a current or a real-time power reading. There's a lot that goes into this.
Mannesson: Yeah, and I think there on that note you're going to see also the processing requirements out in those edge nodes. I was mentioning that in the beginning as well—the edge processing, connectivity aspect of it. We would want to use those solar energy systems, and energy storage systems that we put on the grid to do grid stabilization. That would mean that we will have Linux-based processors—Sitara AM62X-type of capability plus Wi-Fi in those edge nodes so that we can much better manage the energy.
Some customers here probably will want to try to control every aspect of this and have something on their phone where they can decide almost instantly if they want to sell energy, or they want to store energy, or they want to put it inside their car, or what they want to use the energy for.
But there are also going to be consumers that want to just set and forget their system and optimize it on convenience or cost savings. Those are the things that the systems that we and our customers are working on with help with.
DB: So before we get to big picture grid management systems and that sort of thing, I'd love to talk a little more about precision measurements, which is a pet topic of mine, if you can't tell by the oscilloscopes behind me. Let's talk about how precise and how good do these measurements have to be of power consumption in a home or a business.
Mannesson: Accuracy and precision is not maybe always driven by the requirement for billing. I think for the requirement for billing and so on. We have had solutions for a while that is capable of doing this.
I think they are more driven by the requirements, for example, to have a very low THD when you provide power back to the grid. So a lot of the grid operators, they set requirements on how much total harmonic distortion you can have on your 50 or 60 Hz sinusoidal wave that you provide back to the grid. And you need accurate energy measurements to be able to control that in your real-time microcontroller that then is creating that 50 or 60 Hz curve.
Another example would be high bandwidth measurements that can be used to, for example, avoid a DC current build-up in a transformer. So you could have high switching and new bandgap devices like GaN provide that opportunity for us to do very high switching frequency systems. When we do that to measure and to maybe block out a DC current, and to be able to do that, that's going to become very important.
I think TI has a very long history of isolation precision devices, so this could be ADCs or precision amplifiers and so on, that can be used in many of those applications. I think more recently we have also started to release new families that are based on Hall sensors. So that would be TMCS1120xx types of devices; those can be very well used in those applications. When you need to track a fast switching current in an inductor, for example, so that could be used to then implement a very complex control topology. Or you might need a very fast propagation delay, which helps you then to protect a fast switching device like a GaN FET.
DB: So there's a lot more than just figuring how much to charge or how much to pay a person. It's also not necessarily just how many digits of precision or how accurate are we. There are complex measurements that need to be made pretty cheaply.
Mannesson: Yeah, no, exactly. Again, it's not necessarily always the accuracy and the precision. There are a lot of applications as well where, as I said, for example, the bandwidth of the measurement is very important, or the propagation delay. So how fast, how much delay can you actually deliver this value back to a microcontroller that then could take a decision on it. That could be either for control purposes or could be for the purpose of protecting expensive switches.
DB: Yeah, we were talking about safety earlier. Safety for both equipment and people is important. Can you talk about some of the ways that can go wrong? You mentioned like high switching currents and then what TI is doing to help address some of those.
Mannesson: Wwe can take two examples there I think that are common from this market that could be like RCD. So that's called residual current detection. So that's something we do. It's actually fairly similar to a GFCI, so basically a ground fault detection.
We do that in EV chargers to make sure that cables from EV chargers…they definitely face the risk of lying around on the ground, someone maybe drives over them with a car. If you do that a couple of times, that could compromise the insulation of the cable. If that happens, you could start having current leaking out to ground and so on. That could become very dangerous then for any human being that gets close to this EV charger.
So what we do around that is then TI's ability to have both sensing elements—that could be the current sensing element of it, but that's maybe not always enough—you also need to be able to combine that with processing. So that would be the ability to then have real-time processors like a C2000 to be able to do it. Or maybe sometimes low power is very important depending on where exactly you try to do this RCD measurement. So, that combination of sensing element plus processor plus software, then that is a system solution where often my team gets involved in trying to help customers solve.
DB: Yeah, I was just reading into I think it's the CCS protocol for charging EVs. The very first thing it does, once the car is plugged in, is it checks the cable. It does a cable check. It runs the voltage high, and senses to make sure that the cable is good to go. So it makes sense, I didn't dig into the sensing specifics but it makes sense that this would be a big area of focus.
Mannesson: Yeah, no, I mean there's those types of things. There's those things inside the protocol, and it also even happens even if you don't use a protocol like ISO-15118.
DB: We’ll be back with Henrik in a moment. But keynote speakers for Industry Tech Days are sponsored by Mouser Electronics. As you design fulfillment distributor, Mouser is your authorized source for millions of components from thousands of leading brands. Come to discover, design, and develop.
And now, back to Henrik Mannesson, General Manager of Grid Infrastructure and Power Delivery at Texas Instruments.
Real-Time Processing in the Power Industry
You also mentioned real-time processing. I love the phrase real-time. I get the word or phrase real-time because it means so many different things to so many different people. Can you tell me what's driving the need for real-time and what does real-time look like in this context?
Mannesson: I think the key here is that we need to line up production and consumption better in the grid network overall. So the consumption or the generation is fairly predictable, but it varies, and we need that generation capability to be much more in line with our consumption. To be able to do that, I think we need end nodes in the grid network. That could be an energy storage system that then knows the price of energy. Because actually what's going to control this in many times is the cost of energy. So if there is a scarcity of generation, prices of energy are going to go up to encourage people to use less of it.
Then if we have abundance of energy—which we very often have during midday because solar generation systems are providing a lot of energy—and that is then going to encourage people to probably take that energy, store it into an energy storage system, or store it into the car. To be able to do this we probably need information in minute intervals. That's a lot more often than we have it today. Meter networks today, they work often in 15-minute type of intervals. That data is provided hours later to the utilities or for you as a consumer to be able to access and see. From that perspective, if we go from 15-minute intervals down to 1-minute interval, that's a lot more real time I would say. But there are also then use cases (and also coming to people's home) that is going to need much more real-time than that.
There is a concept called frequency containment reserves for disturbances. So this is basically in a grid network today we are trying to keep the 60 Hz or the 50 Hz fundamental frequency of the grid network very, very tightly controlled in very tightly controlled intervals.
Up until today we have had a lot of generators, so large inertia moving around in the grid network, that has helped tremendously with that. With more and more solar generation and much more digitally controlled AC and DC converters, we don't have all of those generators anymore on the grid network.
So to be able to stabilize the fundamental grid frequency, we're going to need to do this a lot more in a digitally controlled way. Here our participation is going to be in this edge processing side where we, on those Linux processors and so on, can run those complex algorithms, use that to either charge energy into an EV or to an energy storage system, or provide a little bit of energy out from those systems.
A grid network or a grid or utility, they're going to be willing to pay consumers energy for having this functionality. At the end of the day, you're not going to provide a lot of energy but that sync/source capability there with a very fast reaction time is going to be worth a lot for a utility. And then we are talking about real-time in seconds of intervals.
DB: Yeah, so the ability for consumers, and maybe a fleet management or an electric vehicle dealership or something like that, to store excess energy as it comes in when it's cheap, but also their ability to help balance the grid system becomes a very vital contribution that they provide to the grid providers.
Earlier you mentioned Linux processors being deployed out into homes and systems. Can you tell me a little more about how these systems are advancing and growing and what sort of new capabilities they're adding on top of existing systems or are they a full swap?
Mannesson: Yeah, no I think you mentioned ISO-15 118 before. So that's the CCS communication protocol between a car and an EV charger. So that protocol is the protocol that's going to help you to do bi-directional charging. It's also the protocol that's going to help you to exchange consumer data in a secure way between the car and the charger.
So this would mean that you could do plug and charge. You could drive up to a charger, plug in your car, the charger is going to recognize the car, and then be able to provide billing information for that charging session.
DB: Yeah, now that you say that, it's kind of bonkers that that doesn't exist already, that you still have to type in your personal information in the app or whatever to make it recognize you in your car.
Mannesson: Yeah, no, exactly. Those things are coming. As I said, it's also what's going to enable bi-directional charge. Because in beta bi-directional charging over that protocol, the car would, for example, know how much energy the battery actually has (that is in the car), and maybe also know how much energy you would want it to have at a certain point in the future.
So that is again something that's going to help stabilize the grid network or help a consumer to then optimize their charging for cost savings or efficiency or flexibility, whatever is most important to them.
DB:If money were no object, and maybe even technological limitations were no object, what would the ideal efficiency system look like to maximize home or business efficiency?
Mannesson: I guess to have solar to cover a large part of your demand, and then combine that with energy storage, and then a little bit dependent on where in the world you are. But probably for a lot of places in the world, heat pumps will make a lot of sense as well.
If you combine those three things or four things together, you have a solar system that you now can choose to either store the energy inside an energy storage system, or you can store it in your car, or you can use a heat pump to generate either heat or cold air with it. You can, in some way, store that inside your house, because you could run that a little bit more when energy is cheap, and then you can run it a little bit less when energy is expensive. Then you are going to be set up for a system that's going to give you a lot of flexibility.
Power at the Grid Level
DB: So if we transition over from kind of home and small business considerations, we've talked a bit about grid overall, but what are some of the biggest differences between small and individual or small business scale systems versus large industrial power systems, both generation and consumption?
Mannesson: A lot of the principles we use are the same. You can have a string inverter that generates 10 kW or 5 kW or you can build a string inverter that generates 150 kW as well. Both of those technologies are definitely there.
A similar thing on ESS systems. There we have battery monitor products that go inside there, and actually it's very often the same type of battery monitor products and battery energy cells that are being used in a large storage system as used in a small storage system. It's just the number of battery modules that you stack up is a lot higher in a grid-scale system.
DB: Where are technological advances happening the fastest? Any specific areas of research or specific industries?
Mannesson: Yeah, I think we talked a little bit about development around wide bandgap devices and GaN and so on. I mentioned it a few times before here, but maybe they'll go in a little bit more detail there. GaN, which TI is heavily investing in, in this space, is something that we see today becoming very prevalent in micro-inverters, power optimizers, and those types of grid systems there. It helps with driving down the cost, making that equipment much, much more efficient and much, much smaller.
So that is definitely a trend that we see. And I said there, I expect there—the home or the commercial part of the grid—to be a faster adopter of those types of technologies versus the more grid-scale type of things.
DB: Can you tell me a little bit about the advantages of a GaN type or a wide bandgap device? Where does that become helpful versus a more traditional device?
Mannesson: Yeah, it really comes down to a couple of different things. First, it's the much higher switching frequency that we can achieve on those devices. This very high switching frequency drives down the size of passive components.
So maybe the largest cost savings and benefits in terms of power density do not come directly from that. The semiconductor becomes smaller. They are small from the beginning. But they come from the fact that we can drive downsize inductors, transformers, and capacitors. So that's one aspect of where GaN becomes important for us.
The other aspect of GaN would also be that we can implement new types of topologies that haven't been possible to do before. We can do them with the help of GaN, and that also helps drive down the size and the cost of some of those grid systems.
DB: I remember the first time I came across someone trying to do a wide bandgap measurement. They were looking at probes or something, and I thought they were kind of joking with the specs they required because the switching speed was so high and the voltage levels were so high. I was like, what are they trying to do that uses this sort of requirement? And it was wide bandgap, and it's been pretty significant technological advancement to see those processes make it out into the wild.
Mannesson: Yeah, that is really what enables a lot of those things that we see today when it comes to AC-to-DC and DC-to-DC conversion on the grid. It's pretty amazing to see some of those advancements that are happening.
I know here in Texas a couple of weeks ago there was a real energy crunch in the early evening hours just as the sun was setting. Then before it starts to cool off, the air conditioners are running very high, and you don't have so much solar energy generation.
There were days there where grid-scale energy storage systems were providing 2-3% of all the energy in the Texas grid system. So, I mean, a short amount of time—of course, it's not throughout the whole day—but during the peak hours there in the early evenings. And that's pretty amazing to see that battery cells, combined then with accurate monitoring and balancing of those cells that is done by small semiconductor devices, and then combine that with large scale AC-to-DC or DC-to-AC converters, that provide that energy out to the grid in that scale.
DB: Can you describe to me some of the differences that might surprise someone who's used to working with small-scale DC-to-AC converters? What surprises folks about a large-scale converter and what sort of special considerations might you have to use when working with one of those?
Mannesson: I think one of the main differences, of course, is that the DC bus voltages in those larger systems tend to be a lot higher. So if you look at home energy storage systems or maybe home solar inverter systems, they tend to be either 400 V or 800 V DC bus voltage. If you go to the grid scale, it's always 1500 V and above.
That level is probably going to keep creeping upwards, because if you want to increase the total power that the systems can generate, we can only increase the current so much. At some point in time, you're going to have to start to crank up the voltage or keep raising the voltage. That is probably the biggest difference.
For us as a semiconductor manufacturer, what that means is, of course, that our investments in things like isolation become important so that we can measure voltage and currents at those very high bus voltages. So that's one thing.
But then, of course, also when it comes to those wide bandgap devices or gate drivers and so on for those devices so that we can control those high voltages. And that is an area where we see those wide bandgap devices come through much earlier than maybe in other places.
A Look Into the Future of Energy
DB: Okay, so if we take a step back and look at the systems overall, what is the state of grid-powered consumption and delivery today, and where could it be going forwards in five or ten years and also maybe even 50 years?
Mannesson: One aspect that we see start to come into this is, of course, machine learning or maybe more popularly called AI. This will help those systems. We talked a little bit about it in the beginning about how you could maybe set up your system to optimize it for different things. Some customers would want to maybe control everything from their phones. Others want to set and forget something, and then just hope this system optimizes for their use case from day to day.
I think that's similar to if you think about how things have evolved in the smart home or building automation space. We came there from maybe the first type of smart thermostats or smart switches and so on. When we started to install them one by one and had many, many different apps to be able to connect and use them, maybe it wasn't that really useful. I think where it really starts to become useful is when you have multiple of those things connected together.
In that language it is probably scenes or something like that. So you could set up different scenes for your home, and that will turn on and off lights,or control your air conditioning system, or start your garden watering system, or whatever that is that you put to it.
I think we're going to see a similar development over probably the next five years when it comes to energy. You're going to be able to either control that directly from your phone, but it could also be that you just let the system (and that's where the machine learning aspect comes in) where you just tell the system to learn from you for a couple of days. When do you usually leave with your car in the morning? So then the system knows—or maybe even knows it from your calendar and your phone and knows where you're going to be at which time—so it already knows how much energy you need in your car and will optimize the charging for that.
Or maybe or maybe you set it to say that you always want to have full flexibility, so you always want the car to be fully charged at some point of time in the morning. You're probably going to have to pay a little bit more for that because you might not use the energy in the most effective way, but it gives you then maximum flexibility.
And I think that's the same as say operating on your building automation or smart home systems. I think you're going to start to see that. Because that's going to be one aspect of how we want to get generation and consumption much more lined up.
The other aspect I think was what we talked a little bit about before, and that's on what type of equipment we will install inside homes. We will probably have heat pumps, energy storage systems, EV chargers, and so on. So that cars, for example, when they are parked, they are connected to an EV charger. Because, if they are, we can actually charge them or discharge them much more effectively than if they are not connected, because then we cannot really use the energy that they have.
So I think if we take those two aspects together, that's going to be very powerful for aligning generation and consumption.
DB: What sort of interesting opportunities and challenges do we face between where we are now and the future vision you've just set for us.
Mannesson: That's where I think Texas Instruments comes into this picture as well. You're going to need a lot more processing. So we talked about the importance of edge processing in this. You're going to have to run a lot more advanced protocols out in those edge nodes of the network. So that's one aspect of it.
We also will have more connectivity, because those different edge nodes need to be much more aware of things that are happening around them. Just as is in the building automation example, if we can connect things together—so an Occupy sensor together with a smart switch—the smart switch is going to become a lot smarter if it knows if there is someone in the room or not.
The same thing in this energy management system. The more those nodes know about things like what the prices are, and what the price tomorrow is going to be, and then also more aware about the user's usage of vehicles, or maybe things like the outside temperature, which then is going to help decide on how to run your heat pump or an air conditioning system.
And of course, for all of this, we're going to need a lot more sensing as well. So that was the aspect of current and voltage sensing that we talked about before.
I think for the high voltage side of things or the high voltage conversion side of things that we talked about; I think what's going to happen there is we will see a lot more DC buses being established, and it starts off in commercial buildings or factories.
So there we have already started to see it happening where you are establishing maybe an 800 or 1200 or 1500 V DC busm which you then can use to run large motors and so on. I think you're going to start to see the same thing happening in a home where you would run heat pumps, cool pumps, air conditioning systems, water heaters, and so on, large appliances from a DC bus that we established.
From a Texas Instruments perspective, we of course then see an increased need there for AC-to-DC and DC-to-DC conversion to move energy around between those different systems. That's where we are focusing our efforts on helping customers do that.
Wi-SUN for Grid Energy Management
DB: One thing that stands out to me…I know TI has been a big supporter and proponent of is the Wi-SUN protocol. In a breath or two, can you tell me about what that is and why it's important?
Mannesson: Yeah, so we talked about those early meter networks that started to come out back 10-15 years ago. The big thing that the Wi-SUN network there provided was a large-scale mesh network. Some of those are probably still in the world; some of the largest connectivity mesh networks that exist.
You can reach millions of nodes in some of those large-scale meter net mesh networks that have been built, with a very limited number of data collectors and repeaters. So basically you relay messages on from meter to meter to get it back to the utility or the energy operator.
Now, I think if you look at that towards the future, what we see there is, of course, that the need for getting data in more real-time is going to increase. So that probably means that the bandwidth of those networks is going to have to increase, and the amount of data that we want to push through those networks will also just continue to increase to create this higher level of awareness across the systems that participate in this grid network about things like pricing, consumption, generation, and so on.
DB: Okay, well, we're about out of time. We like to do a quick lightning round at the end of each of these. So I have a couple of quick questions. How has spending the last blisteringly hot summer in Texas impacted your views on grid management and the responsibility of engineering to make this a more successful and better system?
Mannesson: This is a good question, and we talked about this a bit at the beginning.
The importance of this energy transition that we are in the middle of right now is going to have a huge impact on those things. Basically, what we are trying to change is how we generate energy, how we store energy, and how we consume energy. And that's not a small thing. To try to go and change those things together probably accounts for 30-40% of all the CO2 emissions in the world.
I think if we can get this right with transitioning from burning fossil fuels into generating energy from the sun and through wind, and then be able to store it effectively into batteries, and then also power our transportation with it, that's going to have a large impact.
DB: I feel like changing any of the generation, delivery, storage, or consumption, any one of those four would be a big undertaking, but we're essentially trying to do all of them at once. So it's a very big and very real problem.
DB: You might make some enemies with this one, but what's the best barbecue place in Dallas?
Mannesson: You know, I like hardy barbecue. I've actually come to love brisket, so that's really good there. I've tried a couple of times at home. I'm not really good at it. I've definitely not mastered it, so I'd rather go to a barbecue place for that.
DB: Fair enough. Final question for you. You've spent 18 years at Texas Instruments, something that is a bit unusual to spend so long in a company these days. What has kept you there and what gets you excited and keeps you around at Texas Instruments?
Mannesson: I think I would say the people, first and foremost. You get to work with a lot of very, very talented people here. Those are a lot of engineering and technical skills. I see them in my team every day. I see them in other teams around the world.
And that's probably the second aspect of this, just to be able to work with people, field teams, customers around the world. That's really exciting.
DB: It is really exciting. Thank you for joining me in the Lobby today.
Mannesson: Thank you so much, Daniel. That was good.
DB: That was Henrik Mannesson, General Manager of Grid Infrastructure and Power Delivery at Texas Instruments. Thank you to Mouser Electronics for sponsoring this episode. I think we could have run that keynote all day because there’s just so much to unpack there, but clearly there’s a lot of work to be done and a lot opportunity in the space of power and grid.
But that’s it for today’s episode of Moore’s Lobby. If you liked it, please share this episode with someone else who you think would enjoy it and on social media. I’m Daniel Bogdanoff. Thanks for joining me today.
I’ll see you back here in two weeks when I’ll be joined by Oliver Dial, the CTO of IBM Quantum.