Make Your Own Energy

Efficient use of power and energy in electric vehicles and smart buildings will require innovative thinking.


Regenerative braking and other forms of energy capture are becoming more popular and increasingly effective. What started as a way of increasing the range of electric or hybrid vehicles is now being applied to everything from green buildings to industrial robots.

The automotive industry is still the main driver of this technology. The idea that braking can generate energy has been around for more than a century. It is standard in almost all electric and hybrid cars today, and as the volume of electronic content in vehicles continues to grow, it is finding applications in all types of vehicles. It also is evolving, becoming more sophisticated and more efficient.

“You can adjust the level of regeneration,” said Burkhard Huhnke, vice president of automotive at Synopsys. “This means when you lift your foot from the pedal, the car automatically slows down, and you’re regenerating energy. You can adjust this level because sometimes you just want to just coast before you brake for a traffic light, for instance. It’s the same energy required to accelerate the car to 35 or 50 miles per hour, but usually when you brake all of that energy is lost in the braking system. This can be regenerated by using the motor as a generator and create, for a short time, a current flow back into the battery. That’s a huge advantage of electrical cars because they don’t require an additional system. If you use combustion engines, you don’t have the energy storage, which is the battery package in electric cars. And you don’t have the electric motor, which can be used as a generator if you turn it around the other way.”

With powertrains, the focus is on recovering large quantities of energy, usually from the kinetic energy of the vehicle.

“People have been doing regenerative braking for awhile, and they’re now looking into regenerative suspensions and other technologies of that ilk,” said Jeff Miller, a product marketing manager at Mentor, a Siemens Business. “There’s a lot of work being done in power electronics for big power. Silicon carbide and gallium nitride are being used for some of the higher-frequency operations that lend themselves to driving the powertrain. There’s also work surrounding methods for pulling these odd voltages at odd times and getting it back into the pack in a useful way.”

On the other side, the focus is on powering small devices such as sensors. Various places in the car are inconvenient to run power from the main battery or the accessory battery in electric vehicles, such as tire status monitoring systems.

“The tire pressure monitoring systems of today are all battery-powered, but people looking out to the next generation, especially for things like truck tires and heavy vehicle tires. They are looking at how do we better monitor the lifespan of those tires. Can we put the sensor in the actual rubber that meets the road? For that, performing more kinds of sensing and in even more inhospitable locations, vibrational energy harvesting or other things like that start to make sense.”

Another method of harvesting energy in the automotive space is from vibration of the vehicle suspension, chassis and shock absorbers, which absorb energy resulting in heat. “Here, you lose energy in the range of a couple of hundred watts up to kilowatts,” Huhnke observed. “The trend now is that instead of using a suspension, you can use a piezo element, and if you put tension on to this element it generates electrical charge. A current could then flow back into the battery system so energy could be harvested from the suspension system using this vibration. It’s just physical energy that can be converted into electrical energy. The result of that is not comparable to the regenerative braking because it’s smaller amplitudes, and the overall contribution to your battery charge is limited, but at least it’s something.”

Still, regeneration is an additional opportunity to save a vehicle’s braking system, he pointed out. “You could imagine optimizing your driving by knowing what’s around the corner, and if the traffic light is turning red, then you could adjust your driving style. That would save a lot of energy, as well. In a car-to-infrastructure scenario, you can save a lot of energy by telling the car the traffic light is going to turn red, so slow down, and maybe combine it with the regeneration mode so you can operate most efficiently. Unfortunately, what we realize on our roads today is usually the traffic light race.”

Where it works
Energy harvesting has been important to automotive systems, but not necessarily at the SoC level, said Kurt Shuler, vice president of marketing at Arteris IP. “In EV and hybrid automotive systems, regenerative braking is common and there’s efforts to harvest vibrational energy using piezoelectric transducer MEMS, but this technology will take a while to become mainstream.”

At the SoC level, the first place Arteris IP saw energy harvesting implemented was in 2014 with TI’s SimpleLink CC26xx energy-sipping IoT chips, which are designed to be powered by a separate MEMS-based power source. Even though these chips are relatively simple SoCs from a processing viewpoint, Shuler stressed that they are hugely complex from a power management standpoint. There are more than 20 different power and voltage domains along with dynamic voltage frequency scaling.

But energy regeneration also has other applications. Guy Moxey, senior director of power products at Wolfspeed, said there are two main areas of activity. “One is in the automotive space, where you generate some electricity, and if you don’t need it all you plug the car into the grid and sell excess energy back to the grid during peak-use hours. To do that, you need bi-directionality in the power converter. So that can either be vehicle-to-grid or vehicle-to-home. The second area is where you generate energy more conventionally. So if you think about an elevator in a 60-story building, it’s moving continually and half the cycle is braking. An elevator is typically 20 kilowatts.”

This becomes important as new regulations are applied to commercial and residential buildings. California passed a law in 2018 requiring all new homes to have solar power. In addition, the U.S. Dept. of Energy has developed guidelines for green buildings, as well as a set of recommendations for states to implement. Corporations have adopted these programs, as well. In fact, Synopsys announced last month that its goal is to be carbon-neutral. Adobe has had a similar program since 2010, Microsoft since 2012.

“We’re seeing more of a push to put in LEDs and solar, along with regulations about capturing waste energy,” said Moxey. “And any place where you have energy storage on-site, that’s even better. If you have solar panels, a grid feed and a bank of nine elevators, and you have your own storage, you can capture energy and put it into batteries on-site. We see a lot more of that in Europe. The houses are smaller, which makes it easier to supply the energy needed by people, and most of them have solar. This saves in another way, too, because if you send electricity back to the grid, 40% of that is lost in transmission.”

That’s just scratching the surface of this approach, too. As home or industrial motions move forward or backward, the moment they stop energy can be captured. The difference there is they are not moving as fast, so the braking/stopping time is shorter. As a result, the energy must be captured in seconds or microseconds.

This, in turn, determines the kind of switch that’s required. “You’re either going to use an IGBT or MOSFET,” said Moxey. “The key is the speed and how fast it recovers. So even if you have a diode that is incredibly good, if it recovers slowly that doesn’t work. You also can do things with materials like silicon carbide where the same circuit is used for converting from AC to DC, and then you use a rectifier bridge to convert it from DC to AC. The reason is that with SiC, drain to source is from bottom to top using a body diode. If you do that in silicon, you need separate diodes.”

This also changes the cost equation. While silicon carbide is more expensive per amp, because it can switch faster, the size of the magnetics is smaller and the overall cost of the system is lower.

The efficiency equation
In all of these applications, the key is efficiency. In electric vehicles, that could add significantly to miles per charge.

Zero emission BMW i3 vehicle

“Since charging will take longer than before, it’s really important to have the most driving range as possible,” Huhnke noted. “The most important three vectors for the design of electrical cars are efficiency, efficiency, efficiency, so to optimize the current flow from the battery, let’s say, onto the traction on the road from your drive train, that is the key. Here, Tesla did a great job. The efficiency is extremely high. They have thousands of smaller solutions integrated into the entire engineering design flow, which puts them in a champion’s league in regard to efficiency. But how do you compare current consumption versus fuel consumption? Everybody knows miles per gallon very well. The equivalent for electrical cars is usually kilowatt hours per mile.”

That becomes a key competitive metric for the future. “This can be only done by an optimized engineering design solution along from the battery management system, which includes the logic, the control modules, all the electronics, and help to support the driver with a perfect current flow that they require and demand from the gas pedal,” he said. “This optimization, this design flow, is very important. If I compare cars in the current industry, Tesla won by at least 25% higher efficiency. The key is that they understand the hardware and software stack very well, and that refers at the end to electronics and semiconductors.”

Miller agreed that industry and consumers alike need to start thinking in terms of the kilowatt hours per mile or the miles per kilowatt hour. “Right now, the automotive industry is trying to fold that into equivalent miles per gallon. If you look at an EPA sticker on an electric vehicle, it’s got miles per gallon, which makes no sense except in terms of comparison. The consumer needs to know much this is going to cost to operate. It helps people understand these things intuitively because a lot of people don’t have a good sense of that.”

This comes back to powertrain efficiency and how much can be harvested that would otherwise be wasted mostly as heat in things like brakes and suspension components. “That is a critically important thing as we look to move our infrastructure from gasoline powered to electric,” Miller said. “How much we save in that transition from a real energy standpoint has everything to do with how efficient we can make these vehicles. That said, cars are exploding with sensors, and if you have a thousand sensors on the vehicle and each one has a button cell battery in it, that’s a lot of battery changing. So we need to find better ways to power some of these things that need to be powered.”

This is far from a simple equation, though, Miller said. “How much storage do I need? What is my overall power envelope of the measurement and transmission cycle of these connected sensors? What’s the duty cycle of that? Is it okay if it operates intermittently, or does it need to operate at a very consistent schedule? Working backwards from that, the engineering team can determine what the power system should look like. How much energy do I need to harvest? How much storage do I need to implement? The electronics that go into that, the power management chips that go into those kinds of systems, are increasingly complex and sophisticated. They’re trying to deal with sipping little bits of power here and there and turning it into something that can be usefully be stored on a lithium battery.”

The bigger context
Broadly speaking, energy harvesting and regeneration speaks to a larger and global concern about energy and power utilization and concerns surrounding the ability to serve the immense number of computers, devices and data centers that currently exist. More can be done with the compute resources that are already in place, whether that’s in a data center or a car or a smart building.

“Sending data from a sensor in a field or a factory all the way to a data center in Seattle or somewhere is immensely power hungry, it’s immensely inefficient,” said Chris Shore, director embedded solutions, Automotive and IoT Line of Business at Arm. “Bringing power consumption down is important, but how about not transmitting the data in the first place? Process it where you generated it, and only transmit it when you need to transmit the interesting bit. That goes a huge way towards making the network efficient.”

When it comes to energy harvesting specifically, there is a running joke: “If half the population have all these Internet connected battery powered widgets, the other half of the population are going to be full time employed changing the batteries,” said Shore. “There is some truth in that. The battery changing and battery production becomes a massive problem. The more you can harvest energy from the environment, the less you need to do that. The more maintenance-free devices become, the longer they last, the fewer batteries they use, and so on. Just one example of this is a blood glucose testing device that generates enough power to run the blood test just from physically taking the cap off the device. That runs a generator, generates enough power to take your blood sample and do a test and transmit the result. That’s it. It doesn’t need a battery.”

Whether it is energy harvesting and regeneration in automotive applications, energy efficiency of electric vehicles, or a wider consideration of better energy utilization, much is to be gained from more efficient methodologies and approaches to the design of the semiconductor components that make all of these applications possible. The automotive industry is still at the forefront of discovering and developing these technologies, with many opportunities available for existing and new players in the ecosystem to play an integral role in the future of automotive efficiency.

—Ed Sperling contributed to this report.

Related Stories
Where Is Energy Harvesting?
Sensors that have to be smart enough for networking and security have power conservation all set, following a decade of focused effort by chipmakers.
Electric Cars Gain Traction, But Challenges Remain
30% of cars are expected to be all-electric by 2030, fueled by massive investments in technology.


realjjj says:

The suspension needs to be pre-emptive as opposed to the reactive solutions used today and that limits what can be harvested. The vehicle needs to detect road irregularities for path planning and that data can also be used by the suspension.

Brian engineer says:

Cars don’t generate energy, they consume energy. Recovering a little bit of that back from its absorptive systems (braking, suspension) is a tiny part of the energy that went into accelerating the car and pushing it through the friction from tires and wind resistance. And it doesn’t address the embodied energy that went into building the car in the first place. Ditto for solar panels, energy storage, energy recovery.

Maybe recovering some energy that would be wasted is a good goal. But there’s always diminishing returns to any of these technologies of “regenerating” energy. The 2nd law of thermodynamics never rests.

Leave a Reply

(Note: This name will be displayed publicly)