Engineering at the Crossroads

As the industry pivots to automated/connected vehicles, new research and initiatives are guiding SAE to develop training and education programs to help working engineers and students succeed.

September 1, 2018

Lindsay Brooke

"Never have I seen this industry facing so much potential change,” noted Mike Cohn, a driveline systems integration manager. “The big topics you’ve been writing about in the SAE magazine — electrification, self-driving vehicles — won’t hit the mainstream overnight. But they’ll eventually arrive in decent volume.

As traditional OEMs move quickly to establish AV expertise, Tier 1s are under the most role-change pressure, according to those surveyed. (Image: SAE/STRATEGY ANALYTICS)

“Some engineers,” Cohn told Automotive Engineering, “may regard those technology trends now unfolding as a threat. Others may see them as opportunities. But in either case, they bring new challenges to the vehicle-engineering profession.”

Cohn’s 18-year engineering career could be described as fairly typical. It has included stints at a passenger vehicle OEM, a Tier 2 supplier, and most recently a contract-engineering firm that handles commercial and military vehicle systems development. At every employer he’s felt the non-stop pressure to reduce costs and shorten development cycles. There’s also the constant squeeze to improve product efficiency, meet regulations, and generally optimize capital investments.

OEMs and Tier 1s are trying to consolidate ownership (and control the value network) in ACV technologies. OEMs and tech disruptors will likely own the software and human-machine interface for ACVs, those surveyed believe, while Tier 1s and smaller OEMs will likely struggle with commoditization. Their share of value will flatten as automated vehicles mature commercially. (Image: SAE/STRATEGY ANALYTICS)

“That stuff is the typical yin-and-yang of automotive engineering, and you get used to it,” he explained. “But the autonomy trend in particular is different. It’s unsettling. A flood of new players. Shifting market positions. The growing focus on software. More partnerships and tech collaborations than you can count.

“No one can predict where this is all headed and how soon it’ll get there,” Cohn concluded. “I think there’s uncertainty throughout the industry and the engineering community in nearly all areas right now, maybe due to early hype that self-driving vehicles are right around the corner.”

Giving engineers pause are fickle public attitudes. Consumers this year are showing signs of diminishing enthusiasm regarding self-driving vehicles, after an autonomous Uber test vehicle struck and killed a pedestrian in Tempe, Arizona, last March. Three-quarters of people surveyed by Pew Research in May 2018 said they’d prefer human drivers, even if self-driving cars were readily available. Half of respondents said they’d never want a self-driving car.

The ‘layers’ of emerging technology challenges: Survey participants were asked, ‘What areas are proving most challenging from a technology/ engineering perspective?’ (Image: SAE/STRATEGY ANALYTICS)

The 2018 Cox Automotive ‘Evolution of Mobility’ study also found increased public nervousness: The number of respondents who believe roads would be safer if all vehicles were fully autonomous versus operated by people decreased 18 percentage points since 2016.

While such polling indicates consumer awareness of self-driving vehicles has greatly improved — consumers report that they like many advanced driver-assist (ADAS) technologies — they are not so sure about autonomous vehicles.

Before the Uber crash, companies were racing to get their products to market as quickly as possible. Now some appear to be dialing efforts back. They’re slowing the pace and taking steps to highlight their commitment to safety, before rolling out their vehicles to a skeptical public.

New SAE initiatives

Amid the uncertainty and disruption, SAE International has launched new initiatives to benefit the technical community and other key constituents within the autonomous concept vehicle (ACV) ‘ecosystem,’ as well as to educate the general public. In April 2018 it conducted the first in a series of “demo days” to familiarize the public about automated (SAE Level 4) driving on roads in Tampa, Florida.

Training needs related to ACV development as prioritized by those surveyed. (Image: SAE/STRATEGY ANALYTICS)

The organization followed this with an August announcement that it is developing an automated vehicle safety testing program, the initial framework and safety principles of which are expected to be published later this year.

Intended as a living document that will evolve and be updated to keep pace with technology and regulations, the ACV safety testing program aims to provide a consistent set of metrics with which to evaluate virtual driver systems. SAE will lead development of the testing standards while collaborating with ISO and other organizations to ensure that standards can be globally applicable.

The latest initiative was a survey of global mobility-industry engineers working at vehicle OEMs, suppliers, testing and R&D facilities, academia, tech-solution providers, fleet owners and regulatory agencies. Conducted via 60-minute phone interviews during February-May, 2018, this primary research provided a detailed analysis of the ACV ecosystem. It is being used by SAE primarily to develop a roadmap to address major needs, pain points, and development challenges facing engineers in this rapidly growing mobility space.

The survey analysis also is helping SAE staff to better understand the disruptive forces shaping the direction of the industry.

“The objective of this study was to establish consensus on what the issues are — and where are the opportunities emerging for SAE in this period of disruption,” explained Chris Ambrosio, executive director at Strategy Analytics, which conducted the survey and provided the analysis. He said confirming the issues has set direction for SAE to create products and services that will more effectively serve its customers throughout the industry.

Engineering challenges: Those surveyed were asked how the size and makeup of their engineering teams has changed to meet company goals over the last five years — and how they see their teams, or those of their suppliers, changing to meet technology challenges. (Image: SAE/STRATEGY ANALYTICS)

Noted Frank Menchaca, SAE’s chief product officer: “Strategy Analytics’ research into the impact and trends associated with automated vehicle technology, combined with our continual dialog with customers, users and members, will help guide SAE’s strategy as industry’s and government’s neutral, trusted convener.”

According to Menchaca, the survey research points clearly to the need for SAE to develop the training and education programming that working engineers and students need to succeed — “including in software and environmental sensing technologies,” he noted. The survey also confirmed ongoing input from SAE stakeholders: that a “programmatic approach to bench-marking safety and testing vehicle performance against pre-competitive norms and standards is critical,” said Menchaca.

The accompanying charts and graphics highlight some of the information generated for SAE by the survey.

The engineering universe

Strategy Analytics estimates 291,000 engineers working in the global automotive industry in 2017 — 60 percent of them based in China. India, its engineering ranks growing steadily, will account for 15 percent of the total by 2021.

Engineers working on ACV programs in 2017 accounted for an estimated 12 percent of the total addressable auto-engineering market, which is growing at an annual rate of 25 percent. Nearly half of ACV engineers are located in the Asia-Pacific region. North America accounts for roughly 31 percent of the global total. In Europe, Germany, Israel and the U.K. contribute 20 percent of global ACV engineers, about 7,000 total.

Technology challenges

The survey analysis found tech challenges to be foundational — and most acute in sensors and control algorithms.

There is a growing industry-wide need for training, and universities and other typical sources are lacking in important training areas. Survey respondents also broadly acknowledge the coming need for help in testing and simulation. However, it is too early for ACV standards because so many technologies are in development.

Top 5 tech areas

The top 5 technology areas mentioned by those surveyed as key technologies fundamental to ACV business: 1) sensors/vision systems; 2) control algorithms/path planning/collision avoidance; 3) V2X/connectivity/5G/telematics; 4) fleet management/shared AVs/consumer services, and 5) AI/Deep Learning/Machine Learning.

Engineer insights

Survey participants noted the following trends:

The industry is becoming less dependent on the Tier Ones within the traditional supply chain. Tech suppliers are an increasing source for auto makers via partnership and investment for development. Software (s/w) engineering experts are capturing high value design and customization for auto makers. Automakers are driving out costs with modular architectures (hybrid EV). And companies will continue to need to partner with universities for developing technologies and growing/hiring talent.

Software’s growing role

Importantly, there is a critical shortage of software engineers. Most companies are using acquisitions to add specialized skills in software and algorithm development. Also, the shift toward software-based systems is making agile development methods more important moving forward.

Software’s role will continue to predominate. From a solution vendor: “In the future [designers] will move beyond the design of a 90° turn to many new use cases and conditions. The result will be exponential growth in the number of new algorithms...”

Key engineering priorities

Those surveyed provided the following insights: Currently expertise on integration of ACV solutions

with control systems is lacking. Mechanical expertise will be needed more (“the pendulum of need will swing back”) to enable vehicle and control systems integration. Meanwhile, electrical engineering will continue to be fundamental to all vehicle systems, with steady growth expected to continue. Product cycles will continue to shorten.

And, importantly: those surveyed expect process and cultural changes to occur, to address the “Silicon Valley vs. Detroit mindsets” delta.

Training needed

Main areas of training requested: Systems and Integration; algorithm development; AI and Machine Learning; data integration and data fusion; training on specific protocols including telematics. For mechanical engineers: knowledge on the interactions/interference between sensors and current vehicle systems. Robotics is another major training request topic.

Future priorities

Survey participants were asked to state their level of agreement with the statement: The development of ACV technologies would be accelerated if government would enact regulations tied to safety and testing.

Respondents were generally positive on regulation’s ability to speed up ACV development. However, 43 percent were neutral or disagreed that government has the ability to speed development with regulation. But 80 percent of participants do see regulation playing an important role in driving public adoption of ACVs, especially in the nascent years.