Defying the disruptors and Driving Innovation

April 1, 2016

Lindsay Brooke

Need proof of the auto industry’s crazy pace of change? Look no further than the daily news, where reports of each emergent player — and their impact on what remains of the status quo — seem less and less startling. Consider a vacuum cleaner company that has its eye on more than cleaning carpets.

Dyson, whose efficient high-suction machines created a new technology paradigm, has applied for government funding in its native U.K. and is looking to hire up to 500 engineers, ostensibly to develop electric vehicles. Dyson, which purchased Michigan solid-state battery start-up Sakti3 for $90 million, said it plans to invest nearly $1.5B in battery technology through 2020.

Who would scoff at the prospects of a Dyson EV, given the growing influence of emergent technology players in the automotive sector? OEMs and Tier 1s increasingly need them to help deliver in-vehicle connectivity quickly, at lower investment, and to accelerate autonomous car development. Last year M&A activity by automotive suppliers reached nearly $50 billion, more than three times that of 2014, according to PriceWaterhouseCoopers.

Before GM bought Silicon Valley sensor developer Cruise Automation in early 2016, there were several other significant actions. Toyota hired the entire staff of Jaybridge Robotics, an MIT spinoff in autonomous vehicles, to join its Toyota Research Institute aimed at artificial intelligence, among other technologies. Continental AG bought a 3-D laser-sensor (LiDAR) business from Advanced Scientific Concepts. Last year, Delphi Automotive acquired a stake in Quanergy, a competitor to Advanced Scientific that is developing a low-cost LiDAR, and it purchased Ottomatika, an autonomous-driving software company that was incubated at Carnegie Mellon University.

Meanwhile, collaborations among competitors are becoming vital to share technology and development costs and slash time to market. And engineering services providers are taking on greater responsibilities in everything from software development to combustion system design to multiple-vehicle calibrations.

Clearly, many businesses that may have been in danger of being disrupted are now themselves driving disruption. For insight into how “traditional” OEMs and suppliers are navigating it all, Automotive Engineering spoke with four high-level engineers during the course of our field reporting: Dan Nicholson, Vice President of GM Global Propulsion Systems; Jeff Owens; Chief Technology Officer of Delphi Automotive; Denise Gray, CEO of LG Chem Power Inc., and Dr. Uwe Grebe, Executive VP, Global Business Development, Passenger Cars and Powertrain Systems at engineering services provider AVL. Highlights of their observations follow.

DAN NICHOLSON: Rethinking propulsion

“In the whole discussion of Connected Car, a lot of other areas come to mind before engines, transmissions, and drivelines. We’re in the ‘fourth or fifth row,’ so to speak, but we’re not unimportant. Connected Car impacts design of the hard components very indirectly. But as part of a vehicle’s overall controls strategy, we’ve got to be compatible with GM’s global electrical architectures. So we’re directly impacted in the controls space.

“What will this mean to customers? Connected-car technologies give us greater diagnostic and prognostic capability, in which OBD is a huge area of development. We’re working on expanding prognostics to be able to predict problems before they get to a critical level — where prevention may actually be possible, versus waiting for something to cross a threshold and then service it. It will allow us to do more ‘customized’ 60-point vehicle inspections during normal service visits. We’re looking at taking the engine oil-life example and OBD and expanding it to other areas where we can provide predictive information related to powertrain service to the customer using connected car.

“Having the capability to use the vehicle’s electronic sensors to optimize efficiency of the propulsion system and the vehicle overall presents some opportunities with ‘mild’ electrification. But in the U.S. market one has to be careful; there are some OBD-II design issues in considering those types of technologies. For example, if you use a camera as a sensor input to decide when to do stop-start or when not to do it, it has OBD-II implications. Because now you’ve made it, according to CARB definitions, part of the CO2 compliance system because it affects tailpipe emissions and fuel economy.

“In terms of automated vehicles and autonomy, I see that falling into two ‘buckets.’ One is autonomous cars intended for personal use, in the ownership model. It’s a normal car, a Camaro or Cadillac CTS for example, with the enhancement of being able to drive autonomously — [an SAE] Level 0 car with Level 0 performance expectations. The other ‘bucket’ is cars designed not for personal ownership but as part of a fleet. The kind of vehicles that companies providing transportation services as a business are probably interested in buying. It’s hard to know today whether those cars would be fully electric.

“I expect companies like Lyft and Uber to be early adopters of these technologies because of the duty cycles they have, particularly in cities. It could be a selling point, because the customer buying transportation services has a much different expectation of vehicle performance. Probably zero-to-100 km/h acceleration isn’t an important experience! But those services exist today with conventional propulsion technologies, so it’s not necessary.

“In both cases, we’ll be constantly rethinking propulsion-system requirements — this is a fast-changing and evolving space. A lot of the things that are enablers, such as battery energy density, we’re making good progress on. You can never predict when breakthroughs are going to happen, but it’s not just one breakthrough — it’s incremental breakthroughs that sometimes have a multiplicative effect.”

DENISE GRAY: Lithium battery ‘second phase’

“When you talk about automotive batteries as they relate to vehicle electrification going forward, there are three main pillars: performance, manufacturing, and the ability to support the industry. The industry has entered a ‘second phase’ following the shake-out among battery suppliers and the shake-out in battery and systems designs. The first phase, if you will, was to get product out into the industry and to understand if it’s meeting the customer requirements. And so much learning has occurred in that area during the last eight years.

“This allowed the big battery companies who are ‘in it to win it’ to increase our collaborations with the OEM customers and to learn with them, to make the product better. That learning is continuous regarding the core battery cell technology, battery systems technology, the battery-and-vehicle systems integration, and customer use.

“Battery design and manufacturing now has us into the second-generation of lithium-ion technology for automotive and you see that many of the companies who were emerging five to eight years ago are no longer here today. In that timeframe we’ve seen a pretty amazing reduction in battery price due, in large part I’d say, to the design optimization at the OEMs and from the battery Tier 1s.

“And in this ‘second phase’ of vehicle electrification, there’s an even greater focus on full systems development. My company is not just a battery cell supplier going forward, but part of holistic propulsion-system development, all the way to the vehicle integration level. As you know, LG Electronics is providing major component and subsystems to the Chevrolet Bolt EV. LG Electronics coupled with LG Chem and LG Innotek [the latter develops core material and component technologies including EV components, advanced photonics and solar cells] is a ‘three-legged stool’ that has enabled us to increase our responsiveness to meet GM’s needs to supply full powertrain.

“So there’s an advantage to developing battery technology when you know your own company is also working on propulsion components and power control that could go into the same customer vehicle. I’ve always said that the battery cell chemistry and cell design, and the management of that cell, are closely coupled. If you understand the chemistry, you understand how to control it — and better understand the full system.”

JEFF OWENS: LiDAR under $1000 per vehicle

“LiDAR is going to be one of the active-safety building blocks on the path to a fully automated vehicle. We learned a lot in our coast-to-coast drive in early 2015, with 20 sensors including 6 LiDARs fitted to the car. We found out what radar did well, what [camera] vision did well, what LiDAR did and V2V as well. Looking at optimizing a system, we recognized we could maybe make radar a little less expensive because we know we’re going to have vision there — or maybe make them both less expensive because we know we’ll have LiDAR.

“We got a good feel for the tradeoffs. And one conclusion we came to was we have to have LiDAR in our portfolio for SAE Level 3 and 4 automated driving. I don’t think it’s going to become particularly important for Levels 1 and 2. Part of that is the price point and the value received for it, but LiDAR’s value to a fully automated excursion is tremendous. So we looked at doing it ourselves, and at people who are already pretty far down that road. We found Quanergy and felt they had the best roadmap, the best line-of-sight, to get to a cost-effective solution. Now ‘cost effective’ here is in the eye of the beholder because the comparison is to the $70,000, 64-beam Velodyne. That one’s not only expensive but hard to style on the vehicle!

“Many in the industry are concluding that you need a LiDAR sensor on each of the four corners of the vehicle, with a 120° sweep. The idea is to generate a point-cloud of information around the vehicle. LiDAR is one of the best technologies available because it’s the fastest — it works at the speed of light. It gives you instant data at a distance, and you establish a point cloud around the vehicle. That means if you’re not going to do it on top [top mount on the vehicle] you put one on each corner then coordinate the data extraction and analysis.

“In the case of radar you generate a beam or beams of energy, then you steer the beams electronically by phase shifting. Quanergy has exactly the same concept. You take maybe 8 beams — you don’t need 64 — and make the LiDAR electronically scanned by phase shifting of the optical beams. Of course, first it has to be proven that it has the technology attributes we’re looking for, and second that it’ll go down the cost curve.

“We think Quanergy has a great line-of-sight to something that’ll be around a $200 sensor, in volume. That sounds like a lot because you’ll have four on a vehicle. So $800 per vehicle, but it’s compared with a $5000-$10,000 unit today. We think getting cost under $1000 per vehicle will be the ‘trip point’ where the OEs would say, ‘Okay, that’s a value proposition I can afford because of all the benefits you can get from LiDAR.’

“We think LiDAR will give the vehicle planner and platform team a lot of flexibility regarding their choices for the sensor package. The more you want the car to be automated, or take the driver out of the seat, the more you’ve got to ensure the vehicle is very sure of its surroundings. Radar is good at this; radar-plus-vision is better; radar-plus-vision-plus-V2V is better yet — and all of that plus LiDAR is best yet. You also need it for high-def map matching.

“Take the driver out of the seat and you’ve got to have 100% positional integrity, all the time. The car’s performance has to be far greater than today’s if you’re going to hand over control to the computer.

“There’s opportunity for greater sensor fusion as we incorporate LiDARs, as we did with the Delphi Raycam on the Volvo XC90 — it’s the best example of radar and vision fusion in the business. Because of the corner coverage with the LiDAR, you probably won’t combine it with a vision system. But there is a possibility of fusing LiDAR and radar. There’s also the opportunity to replace the vehicle’s ultrasonic sensors. As radar technology comes down in cost, could we put a radar-LiDAR sensor at each corner and share the processing? I’m sure there’s an opportunity there — but its time will come.”

UWE GREBE: Engineering ‘from the virtual to the road’

“The need to develop electrified vehicles, connected-car technologies and automated and autonomous driving systems — on top of the ongoing pressures of faster development cycles and reduced cost — have created a new business model in this industry. And that model is where the OEM takes much more ownership of delivering a great vehicle, but doesn’t necessarily do the engineering of certain parts and systems. And the Tier 1s have more responsibility to deliver increasingly sophisticated, tailored systems.

“This is a new era that is about ‘requirement engineering’ and managing the result of the engineering to where the OEM is confident of the excellence of the final product.

“There is additional complexity that comes with the controls, data processing, and power requirements of the new electrified and electronic systems, of course. As an engineering services provider, we at AVL are in a position of helping to integrate these things with the OEM. We do a lot of bench-marking and it goes way beyond the classical definition, where I tell you, for example, what the part weighs and how it was cost-engineered. Instead, our benchmarking goes deep into vehicle attributes: what is the set of competitors, what is the objective performance, what the customer perceives as tip-in performance, brake-pedal, whatever.

“This helps the OEM define the market requirements for the Tier 1 and engineering service provider for development. So the OEM can say, ‘I follow your release recommendation. I would have done it in the same way if I’d had my own people doing it.’

“This is a new scenario for AVL and other engineering service companies. But there is also a growing demand for developing far-reaching ideas or implementing a new technology. And it’s still an important part of AVL’s business, by the way, particularly in the powertrain arena, which is a core strength. The innovative stuff, providing a deep, deep understanding of combustion systems, for example, and helping OEMs through their technology roadblocks.

“If you look at where the OEMs are investing in engineering resources, it’s on the software and algorithm side to support the safety systems, infotainment and connected-car areas. It’s not so much in traditional mechanical engineering, but there is increasing demand for more efficient and reduced-emissions engines and their drivetrains.

“To support work in these areas, we have new developments in what we call ‘from the virtual to the road.’ Our methodology is called the Integrated Open Development Platform. Think of the office environment with the math tools then going to Software in the Loop, then hardware in the loop, then to a powertrain dyno, then to a chassis dyno and on to the road.

“What is important here is consistency of test cases and models. You can front-load into the math environment, but you need to be sure that whatever the model is for this powertrain or vehicle, it must have fidelity and needs to be representative. You have to ensure the vehicle performs in the same way, and your math model has to do the same thing. So these are the test cases you put through from the real-vehicle environment to the math environment.

“It’s called ‘Integrated and Open’ because every customer is free to choose. Our customers like this approach, even though we like for them to use our tools! It’s always difficult to accurately look into the crystal ball. I see continued growth for AVL and for our part of the business. I hope for many more partnerships in the industry.

“Automotive is a highly competitive environment and that’s a good thing, it keeps you running fast. But in this triangle of OEM, Tier 1 and engineering services provider, partnerships are crucial going forward and we need to maintain partnership continuity.”

---