Follow-up report, 16th International CTI Symposium – Automotive Transmissions, HEV and EV Drives
4 – 7 December 2017, Berlin, Germany
A plea for technology-neutral thinking
Increasingly, drive development work calls for ‘hybrid’ developer qualifications that include both traditional mechanical thinking, and new methods from Big Data, Artificial Intelligence or machine learning. Two other factors are also growing in importance: an open mind for technology-neutral thinking, and a comprehensive view of where CO2 emissions actually occur during vehicle lifecycles.
Symposium chair Prof. Ferit Küçükay opened the 16th CTI Symposium on 4 December 2017 by reviewing just how diverse topics have become in Berlin, as well as at the symposia in the USA and China. Naturally it’s all about transmissions and components such as start-up and shift elements, actuation and lubricants, together with various levels of hybrid drives, add-on solutions and DHTs. But increasingly, other topics are coming too. These include machine-based learning, artificial intelligence and data fusion in the cloud, with all the opportunities these offer from production right through to the drives themselves. All these topics were addressed at the symposium, which was attended by 1,350 participants. Together with the symposia in the USA and China, that means more than 2,700 experts from the drive developer community got together in 2017. In Berlin, attendees could choose from more than 100 lectures and visit 120 exhibitors. The topic bandwidth was also reflected in the plenary talks, which were as diverse as the challenges that drive and transmission developers will face in years to come.
In the factory of the future, the only limits are the walls
Rolf Najork, CEO Bosch Rexroth, has been pressing forward with networked production solutions since 2016, based on the insight that drive electrification makes drive concepts more diverse, and hence demands more flexibility in production. Mr Najork explained that more individual ‘small batch’ production needs shorter make-ready times, lower investment costs, and to a certain extent new manufacturing procedures such as 3D printing. He described the ‘Factory of the Future’ as a completely variable space where the only limits are the walls. The production facilities inside are mobile, can be swapped out quickly, and induct their energy from the floor. Mr Najork said Rexroth has already installed this type of ‘autonomous space’ at one facility. He noted that the Factory of the Future needs to be cable-free, adding that you need a standardized communication protocol such as 5G to control a digitalized value chain. In his words ‘the customer designs the car’, which ultimately means digital communication is required from product definition right through into production. The speaker believes tomorrow’s customers will pay for ‘uptime as a service’ with faster availability meaning higher profits, and foresees more leasing models for flexible production facilities. Mr Najork said China and the USA currently lead the field in digitalization overall, but Europe could be a trendsetter in hybrid fields, where mechanical aspects are still important.
Efficient drives need a Cradle-to-Grave assessment
Prof. Dr Peter Gutzmer, Deputy CEO and CTO Schaeffler, and Dr Wolfgang Warnecke, Chief Scientist Mobility Shell, dived deep into the pros and cons of various energy sources. From a cradle-to-grave viewpoint, they said, there is little difference between diesel (156 g/km) and EV (Tesla S: 135 g/km, of which 65 g/km is just battery production alone). They also pointed out big differences between e-fuels, hydrogen and batteries, noting that the energy density of petrol is around 32 MJ/kg, compared to roughly 8 MJ/kg for hydrogen (including the tank system) and less than 2 MJ/kg for battery. On the other hand, Warnecke said, hydrogen and e-fuels perform poorly in well-to-wheel assessments. He gave efficiency figures of 72% for BEV, 37% for FCEV and between 12 % (Power-to-Liquid) and 17 % for Power-to-Gas. Gutzmer added the volatile nature of renewables to the equation, saying it remained to be seen how the various energy sources could be stored and transported in the future. Even beyond 2030, he predicts, around 30 percent of the world’s vehicles will still have an ICE combustion engine – 70 percent if you include hybrids. Gutzmer believes many different roads lead to clean drives, saying we need to think in terms of systems and scenarios. He expressed concern that in striving for technology-neutral solutions, legislators occasionally ‘presuppose the existence of technical robustness’.
From automobiles to mobility tools
It’s long been suspected that younger people are less interested in cars these days. Dr Sven Beiker, Managing Director Silicon Valley Mobility, showed that since the mid 1990s, driving distances in the USA have no longer grown in line with GDP growth. Also, just 77 percent of people aged 20 to 24 took their driving test in 2014, compared to 92 percent in 1983. Dr Beiker interviewed Stanford University students to find out what today’s young people expect from automobiles. Surprisingly, he discovered that automobiles may still be ‘emotional’; some interviewees put this down to the fun of driving, others said even autonomous driving could be emotional because it frees up time for other activities. All students expected adequate range from an EV and said the choice between ownership and car / ride sharing depended on the context. As more people share, Dr Beiker sees opportunities for simplifying automobiles and making them around 35 percent cheaper. He said customization could then be achieved through aspects such as personalized infotainment offerings. Most students didn’t care what kind of drive cars used, saying what mattered was range, reliability, convenience and ease of operation. Interestingly, sustainability was hardly mentioned at all; above all, the students saw automobiles as a mobility tool.
Well-to-Wheel assessments for Europe
Allessandro Coda is CTO of the European Association of Automotive Suppliers CLEPA, which represents suppliers and mediates between the industry and legislators. In his talk, he described the advantages of well-to-wheel (WtW) assessments in Europe. At present, fleet consumption figures are only based on a tank-to-wheel model, not well-to-tank or ‘well-to-plug’. This means they completely ignore the CO2 footprint for generating fuel or electricity. Switching to WtW would mean slightly higher CO2 figures for ICEs. The same would apply to electric drives, whereby that figure would depend on how much CO2 was involved in generating their electricity. Hence, electrified drives would have a growing advantage as more and more energy came from renewables. Mr Coda said there wasn’t enough data yet for complete lifecycle (LCA) or cradle-to-grave analyses, adding that these would be far more comprehensive since they would include the entire CO2 footprint for battery production, and disposal later on. Mr Coda sees WtW as a first good first step towards a complete LCA. As questions from the audience showed, however, one aspect calls for further discussion: while WtW includes the CO2 footprint for generating electricity (the speaker expects just 49 percent to come from renewables by 2030), it ignores the very high CO2 footprint from battery production.
How does machine learning work at Google?
Digitalization and artificial intelligence (AI) are becoming more important in the auto industry too. Marc Ritter, Google Manager EMEA Automotive Partnerships, spoke on machine learning, saying that just as humans learn to speak through trial and error, machines are made to keep practising too and are allowed to make mistakes as they learn. The prerequisite for that, he said, is Big Data – or more precisely just data, because you need a lot of data per se. This data is stored in the cloud and is processed by ultra-fast processors that Google has developed in-house. The software used is Google Tensorflow, an open source system for machine learning and AI. Mr Ritter gave two examples of how Tensorflow works. First he described a Japanese family that grows gherkins and needs to sort them by quality for the Japanese market. The son used Tensorflow to construct a simple mechanical apparatus that recognises gherkins by their shape and sorts them automatically, steered by Tensorflow. His second example involved distinguishing between snow and clouds on satellite images: Google Maps removes clouds because unlike mountain snow, they are transitory. Naturally, the speaker said, Tensorflow is available for projects in the automobile industry too. Although he left potential application fields open, obvious examples would certainly include pattern detection for driver assistance systems and automated driving.
Drives 2030: 70 percent combustion engines, 70 percent electrification
Uwe Wagner, Schaeffler, also thinks it would be better to measure CO2 footprints from well to wheel, or better still from cradle to grave. But in his talk, he focussed on real-life solutions for tomorrow’s drives. Mr Wagner assumes that in 2030, 30 percent of the 120 million automobiles produced worldwide will be BEVs and 40 percent HEVs, with the remaining 30 percent still powered solely by their ICE. He said this mix would enable fleet CO2 emission figures of less than 60 g/km, but further improvements to combustion engines were still needed. These could include improved bearings, optimized front end accessory drive (FEAD) or optimized thermal management. More variable valve train was also needed, while on the other hand, ICE phlegmatization was likely with DHTs. Mr Wagner presented interesting figures on potential savings for 48V hybrid drives: 3.8 % for P0, up to 6.9 % for P1 and 14.8 % for P2 (and each another 1 % better with optimized FEAD). P3 would save up to 15,3 %, P4 up to 15.5 %, and even 24 % compared to a mechanical all-wheel drive. Mr Wagner said P0 systems call for in-depth understanding at system level, whether in suspension, belt tension or E-clutches. To master all this diversity, Schaeffler adopts modular approaches – for example E-axles with different ratios. The speaker says Schaeffler hasn’t written off wheel hub motors yet either: for city applications, these have the advantage of not intruding into the passenger cell.
Growing range means bigger challenges for electrified trucks
Anders Nielsen, CTO Volkswagen Truck & Bus, looked at the challenges posed by the electrification of commercial vehicles. He said fuel accounts for about one-third of the costs for long haul trucks, far more than for passenger cars. However, truck consumption is comparatively low: it dropped by 50 percent from 1970 and 2000 thanks to technical improvements in drives, aerodynamics, tractive resistance etc. Mr Nielsen said that figure would halve again by 2020 due to factors such as transport logistics or assistance systems. By 2030, the speaker expects a further 50 percent reduction from automation, connectivity and electrification. For pure electric trucks, he said, infrastructure posed greater challenges because charging times and driver rest periods ‘do not match’, meaning truck stops would need to change completely. Mr Nielsen said diesel’s energy density is roughly 25 times higher than that of a battery. Asked about the recently announced Tesla truck, he estimated that the batteries would weigh 5 to 10 tons. He sees e-fuels and e-highways as potential options for long-range haulage, noting that overhead lines have the great advantage of being 75 percent efficient. For the speaker, the strengths of BEV lie mainly in urban delivery scenarios. Overall, Mr Nielsen expects trucks will keep the same complete portfolio of conventional, hybrid and electric drives as passenger automobiles. For EVs he prefers modular battery concepts as a means of scaling range to application.
Great Wall pursues in-house development of dual clutch transmissions
As the talk by Lipeng Zheng, Vice President, Great Wall Motors (GWM) showed, the Chinese manufacturer covers the entire spectrum of transmission development and manufacturing in its home country. Mr Zheng presented the 7DCT450, a seven-speed DCT with wet clutches and a ‘dual pump system’ that comprises a mechanical and an electro-hydraulic oil pump. He said this pump system alone could reduce fuel consumption by two to three percent, adding that the electro-hydraulic component was an advantage in hybrid drives, but also enabled the oil flow to be toggled between actuation and lubrication. Mr Zheng also cited a list of state-of-the-art activities including the use of low viscosity oil, optimal NVH performance due to optimized wheel sets and a tailor-made shift strategy, friction-reduced bearings etc. He also emphasized the specific requirements of the Chinese market, noting that in terms of robustness and comfort, expectations had risen sharply in a few short years and were now on par with Europe. Mr Zheng also underscored the high levels of in-house manufacturing and manufacturing at his company. While the supplier list includes many established names from Europe and North America, high production volumes mean GWM strives to keep as much of the value creation process in house as possible. That includes all software development work and the testing of over 1,000 automobiles on 26 different tracks.
Affordable hybrid drives are essential for success in the market
Tatsuya Osone, Vice President Advanced Technology Development Jatco, spoke of the advantages of CVTs for hybrid drives, and concepts for PHEVs and BEVs. He sees two categories: firstly, mild and full hybrids that build on conventional drives and remain affordable and secondly, plug-in hybrids and pure electric drives whose price points remain challenging. Mr Osone said the latter category was needed in order to benefit from ‘credits’ in certain regions, or in the case of pure electric even to be allowed to drive at all. Mr Osone still thinks CVTs are particularly suitable for 48V hybrids, where the combustion engine handles a relatively high share of propulsion. Here, CVTs could play out their strengths by keeping the ICE in its efficiency ‘sweet spot’ under partial load. The speaker said e-motors provide around 40 percent of hybrid propulsion, 55 percent in full hybrids and around 80 percent in plug-in hybrids. He believes it makes sense to use an easily scalable ‘building block’ system of e-motors across all levels of electrification. In the second category – PHEV and EV –Mr Osone favours e-motors with a single-speed or simple multispeed transmission that enables a compact build size. He is sceptical of DHTs in this context, saying low costs were essential in order to offer low cost PHEV and BEV solutions too.
Dual track drive development and 48V BEVs
Mahle is strongly linked to ICE systems and components, so it was particularly intriguing to hear what CEO Wolf-Henning Scheider would say about electric drives in Berlin. The company is pursuing multiple strategies, as demonstrated by the Range Extender Drive it first presented in 2012. This has a range of 500 km (of which 70 km are pure electric) and a CO2 footprint of 42 g/km. Mahle also offers thermal management and A/C solutions for electrified automobiles; in this context, Mr Scheider mentioned the company’s partnership with Faurecia to develop needs-based solutions for passenger cell A/C solutions. The speaker said the biggest hurdle for successful e-mobility is range. He also pointed out that a BEV’s lifecycle CO2 footprint is four times that of conventional vehicles, meaning EVs would have to drive 15,000 km a year for eight years to break even, while CNG-powered automobiles would never break even at all. Mahle is strongly in favour of 48V concepts. Mr Scheider presented an e-motor with a peak rating of 30 kW and a 20 kW continuous output that can save around 25 percent compared to a high voltage system. In response to a question, he said the currents involved (up to 800 amperes) were not a problem to manage. The demo vehicle ‘Meet’ combines two e-motors on the rear axle, each currently with a peak output of 20 kW.
Podium discussion: What will drives look like in 15 years’ time?
In line with the plenum talks, the topic of the podium discussion was ‘How will future mobility requirements change powertrains and transmissions?’ The participants were Peter Gutzmer, Schaeffler; Sven Beiker, Silicon Valley Mobility; Uwe Keller, Daimler; and Roland Werner, Uber Germany. Presenter Ulrich Walter began by asking the audience what subject they thought tomorrow’s students should choose. The majority said mechatronics, while many others opted for software and electronics. The results of the audience poll were projected onto a screen in the form of a Keyword Cloud. The panel agreed with the audience: Prof. Gutzmer believes technology neutrality is important in order to master tomorrow’s challenges, and that core technologies should not be challenged. Sven Beiker said mechanical engineering would be needed in future, to which Prof. Gutzmer added that we need a combination of both. With a view to Silicon Valley, which is almost synonymous with tomorrow’s technologies, Mr Werner referenced the term ‘Internet of things’ and pointed out: ‘They have the internet, we have the things’.
The question of which type of drive will prevail is still as fascinating as ever. Forty percent of the audience said hybrid drives would dominate in 15 years’ time, almost 30 percent voted for BEVs. Astonishingly, over 10 percent voted for fuel cell automobiles, while nearly ten percent can imagine not having a car at all. However, those figures shifted noticeably when filtered by age range. In the <40 age group, one person in five could imaging not owning an automobile anymore; in the >40 age group, a significantly higher number (almost 13 percent) think FCEVs will succeed. In the long term, Prof. Gutzmer also sees hydrogen as a very promising option. Uwe Keller cited the Fuel Cell Range Extender as a hybrid compromise. He reasoned that a small battery is enough for daily trips and would charge faster; drivers would only need the modest hydrogen filling station infrastructure on longer journeys. Everyone agreed that it’s important to stay open for competition between different technologies, and to not insist on individual solutions.
Which brings us to the question: ‘Which transmission type will have disappeared from the global market in 15 years’ time?’ In reply, 35 percent of the audience said ‘manual shifts’, while a similar number said no transmission types would disappear. Of the other transmission types, automated manuals were named most frequently (17 percent). Opinions in the plenum were markedly different: Sven Beiker said manual transmissions would still be needed in some regions, pointing out that ‘not everywhere is like Palo Alto and Berlin’. He believes it makes sense to have three modular transmission kits, meaning you can add automation and electrification. Uwe Keller thinks the best solution in 15 years’ time will be compact hybrids in the form of DHTs, saying this was where investments were currently being made.
Every one a winner – 9th Young Drive Experts Award
Following last year’s positive results, three young engineers stepped up once again this year to hold short ‘Pecha Kucha’ presentations. These consist of ten slides that switch automatically every 20 seconds; the winner is the candidate who gets the most applause. In a very close race, this year’s winner by a whisker was Nils Bothen, Leibniz University Hanover, with his Master’s thesis ‘Optimizing a Rotation Transformer with an Electric Steel Core’. The runners-up were Philippe Jardin, TU Darmstadt, with ‘Reconstructing Load Conditions to Calculate Potential Failure Rates of Automotive Vehicle Components’ and Patrick Weißensteiner, Virtual Vehicle Research Center Graz, with ‘Driving Dynamics Configuration of an All-Wheel-Drive Electric Automobile with Omnidirectional Steering’. All three candidates left a lasting impression on the audience by presenting their research findings so concisely in such a short time. In that sense, all three were winners.
A symbiosis of proven and new technologies
What were the important insights from the 16th CTI Symposium? One was certainly the fact that transmission development in China is now on the same level as elsewhere, and seems likely to remain so. That applies not just to highly electrified drives, where China is obliged to act in view of local emissions, but also for conventional drives like the ones based on the new DCT from Great Wall.
The topics of well-to-wheel and cradle-to-grave were addressed repeatedly, and with emphasis. BEV CO2 emissions look fundamentally different when viewed from tank-to-wheel (today’s model), well-to-wheel (the planned model) or cradle-to-grave. Some speakers said the last option would be fair as a way of considering electrification without prejudice. At present, the environmental costs of batteries are only factored into CO2 footprints in a truncated form. This issue will likely keep the industry increasingly busy in years to come, and hopefully the general public will take a more differentiated view too.
Artificial intelligence is also playing an increasing role, whereby what that means in an automotive context is yet to be defined. For the moment, Google’s Marc Ritter left it at machine learning, which basically means enhanced image and pattern detection, but not the ability to reflect on what is detected. When errors occur, the ‘ogre of product liability’ (Sven Beiker) looms, which is why he believes more in a symbiosis of AI and control system technology.
And what will become of transmissions? Today’s diversity will probably still be there in 15 years’ time. There will be more electrification, whether in the form of modular add-on solutions or DHTs. Transmissions will be more strongly integrated in ‘smart’ systems, and software and ‘system thinking’ will play a bigger role. But we’ll still need simpler solutions too, because to repeat the words of Sven Beiker: ‘Not everywhere is like Palo Alto and Berlin’.