Navigant Research Blog

Success in Automated Vehicles Depends on Tech, Services, and Manufacturing

— April 3, 2017

An old axiom in motorsports goes: “to finish first, first you must finish.” This means you can have the fastest car on the track and qualify on the pole position, but if you don’t have the preparation or team to back you up, the quality of the car is meaningless. In the race to make automated driving a successful commercial reality, hype may get companies all the media attention, but a fully realized strategy combined with the ability to execute are the keys to success. This is why Ford, General Motors (GM), the Renault-Nissan Alliance, and Daimler are the leaders in the latest Navigant Research Leaderboard Report: Automated Driving Systems.

Outside observers would not be faulted for believing that companies in Silicon Valley were about to roll over the entire automotive industry and take over personal mobility in the coming months based on news coverage. However, as many veterans of the technology industry have become painfully aware of, the reality is that building vehicles to safely transport the world’s population is far more difficult than just writing an app and publishing it to an online store.

Horse Before the Cart

Assembling a suite of sensors and writing the basic software to control a vehicle are actually the easy parts. Before that package can become a real product, you need a vehicle. Google developed its automated driving system in 2009 by hiring many of the top brains from Stanford, Carnegie Mellon, and several automakers that had previously created winning vehicles in the DARPA Grand Challenge program between 2004 and 2007. Then Google went to local Toyota and Lexus dealers and bought vehicles one or two at a time. Companies like Cruise Automation and Uber followed similar paths. In order to commercialize a system, they will need to invest billions more to develop and manufacture vehicles or find an automaker partner willing to supply cars.

Uber is reported to have lost more than $3 billion in 2016 without capital investment in vehicles or manufacturing. The world’s major automakers already have the engineering and manufacturing infrastructure in place, and many of them have been working on autonomous technology for far longer than Silicon Valley. Major automakers understand the intricacies of developing, validating, and certifying vehicles for profitable production.

At Navigant Research, we believe the leading automakers are learning what it takes to develop automated vehicles faster than new entrants can learn how to build cars. Companies like Ford, GM, Nissan, and Daimler also understand the regulatory and product liability hurdles faced by bringing automation to the world’s roads. These companies have heavily invested in controlling and understanding the key technologies required to make vehicles and automated driving system work seamlessly.

Just Around the Corner

The leading companies in this field are also rapidly developing their own in-house mobility services so that they can provide consumer access to automated driving systems while retaining control of vehicle manufacturing. This will help to ensure that the vehicles are properly maintained and updated—something that is key to safe and proper use within the early years of deployment.

Technology companies like Waymo and nuTonomy, as well as suppliers like Delphi and ZF, will have an important role to play in the new mobility ecosystem. But for now, automakers lead in the automated driving system race.

 

Questions Aplenty About Interacting with Automated Vehicles

— March 31, 2017

Over the past 130 years, the interface between human and machine has become relatively standardized. We have steering wheels, pedals, seats, mirrors, and other major controls in roughly the same location no matter what brand or type of vehicle we use. We’ve made adaptations for additional hand controls for those that have physical disabilities, but overall, the experience is consistent. But when those controls are eliminated in automated vehicles (AVs), as Ford will do in the AV it intends to produce in 2021, designers have opportunities to rethink vehicle cabins. And those opportunities raise a few questions about interacting with AVs.

User Experience

The more you know, the more you realize how much you don’t know. As we accelerate toward an era where humans are no longer in direct control of the vehicles we move around in, it’s clear that making a car drive itself is only the beginning of the task at hand. A panel at the recent Automotive Megatrends Autonomous Car conference in Detroit examined some of the questions around the user experience (UX) with automation.

A crucial aspect of the human-machine interface are the seats and what we see as we move through the world. Over the last several years, automakers and suppliers revealed a number of fascinating concepts for cars of the future such as the Mercedes-Benz F015, Nissan IDS, and the BMW HoloActive Touch.

One of the seemingly more appealing ideas about not having to drive is that vehicle occupants could be repositioned so they can interface with each other instead of the vehicle. However, a driving force behind the design of modern vehicles is the need to protect occupants in the event of a crash. They must be properly positioned in order for airbags to provide protection. While AVs are likely to cause far fewer crashes, they will still have to coexist with the more than 1.2 billion vehicles on the road today and for decades to come. That means that unless we ban human-driven vehicles, AVs still have to conform to the same safety standards and seat rotation will be limited to small angles.

Then there is the whole issue of motion sickness. Many people experience physical symptoms when there is a disconnect between what their eyes see and their body feels during motion. If we go from driving to watching or reading during our commutes, this could become a design issue.

Voice Recognition Systems

Another question regarding interacting with AVs: Will we let self-driving vehicles know where we want to go? For all the attention that devices like Amazon’s Echo have received in the past couple of years, voice recognition systems remain frustrating to use. Companies like Google and Nuance have made huge strides in improving the reliability of these systems when they are connected to the cloud, but even the most advanced machine learning systems continue to struggle with natural language semantics and accents. There is an enormous difference between recognizing individual spoken words and the meaning that is imparted by stringing a series of words together.

Humans are remarkably adaptable, and we will likely adjust our own speech patterns to the limitations of the machines before the machines themselves can reliably understand us. Or we may decide that if technology can’t make our lives less frustrating, we may reject it.

 

Ford Makes $1 Billion Bet on Artificial Intelligence Startup as Recruiting Tool

— February 10, 2017

Electric Vehicle 2Over the past 4 years, Ford has made a concerted effort to move from also-ran status to vying for the lead in the race to produce automated vehicles. The latest move by the company is a 5-year, $1 billion investment in Pittsburgh-based Argo AI. The artificial intelligence startup will operate semi-independently while developing the virtual driver platform for the automated vehicles that Ford has promised to start producing by 2021. However, the deal seems less about acquiring technology today than acquiring new talent in the coming months and years.

Startup Power

Argo AI was founded in late 2016 by Google self-driving car program veteran Brian Salesky and Peter Rander, who led Uber’s automated car program until September 2016. The total investment of $1 billion will be parceled out over the next 5 years to fund the operating expenses of the company and to provide equity incentives to new employees in order to help with recruiting.

Ford plans to retain control of all hardware product development and manufacturing internally, as that’s where the automaker’s expertise lies. The Argo team, which has plans to grow to a staff of 200 engineers by the end of 2017, will work on developing the so-called virtual driver software platform that will control the vehicle. The software team will be integrated into the product development process as it moves forward with trying to make the system robust enough to be able to fully operate without a human driver. Ford has been developing its automated driving platform over the past several years to get its software platform where it is today.

The fact that Ford has gotten this far down the development timeline with a committed production date only about 4 years away does raise some questions. An investment of this size into a startup at this stage implies that there may be issues with the automaker’s in-house software platform. However, Ford chief technology officer Raj Nair emphasized during a conference call to announce the deal that Argo’s expertise will be used to further enhance what Ford has already built.

Attracting Talent

The existing engineers working on Ford’s platform will move over to the Argo AI team to become core employees of that company. As such, they will be eligible for the stock compensation plan that Argo is putting in place, which is typical of Silicon Valley startups. Since Argo has only existed for a few months, it’s unlikely that the company has built a complete system that would replace what Ford already has. Instead, it appears that Ford saw an opportunity with Argo to put in place a corporate structure that would enable it to tackle one of the thorniest issues that the auto industry faces in the race to deploy advanced technology: recruiting.

Traditional companies have long had difficulty attracting some of the top talent away from Silicon Valley, where they see startups as a potential pathway to a huge payday with an initial public offering. Many of those companies with limited funds pay employees with stock options that can be worth millions if the company succeeds and the employee sticks around. That doesn’t happen with regular Ford employees.

Ford CEO Mark Fields acknowledged on the call that an Argo IPO at some stage is a possibility if the company succeeds. Investing in what is essentially a brand new company with some proven leaders in the form of Rander and Salesky is an interesting new approach to the talent acquisition problem.

 

The Future of Smart Parking Is Integration with Automated Technology

— January 26, 2017

Electric Vehicle 2The smart parking industry continues to evolve as an increasing number of cities struggle with traffic congestion and inadequate parking availability. While the deployment of sensor technologies continues to be core to the development of smart parking, a wide variety of other technology innovations are also enabling more adaptable systems—including cameras, wireless communications, data analytics, induction loops, smart parking meters, and advanced algorithms.

Moving Toward Automation

The future of the smart parking market is expected to be significantly influenced by the arrival of automated vehicles (AVs). Several cities around the world are already beginning to trial self-parking vehicles, specialized AV parking lots, and robotic parking valets.

For example, in Boulder, Colorado, ParkPlus is working on deploying the first fully automated parking garage in the Western United States through Boulder’s PearlWest mixed-use development. The company’s automated parking system uses lasers to scan cars and a robotic valet to park the vehicles. Vehicles are transported by a robotic dolly that lifts and transfers them to storage racks. Using this system, up to 4 times as many cars can be parked in the same amount of space as a traditional garage (since there is no need for extra space in between cars). The automated system is expected to deliver vehicles within 2 minutes of a retrieval request.

Development and Deployment

Somerville, Massachusetts has partnered with global automaker Audi to develop self-driving and self-parking cars. In 2018, a small fleet of cars with piloted parking technology will be deployed to test self-parking capabilities with a specialized nearby parking garage. The idea is that having cars that self-park will help improve traffic congestion considerably as riders could be dropped off in front of their destination and the car would park itself and minimize the time spent taking up space on the road (as opposed to drivers circling several blocks continuously looking for a space). In 2020, phase two of the project is expected to commence with the deployment of a full fleet of self-parking Audi cars. By 2030, the self-parking garage is targeting availability to the broader AV market. It is estimated that parking garages specifically designed for self-parking cars can take up 60% less space than traditional lots (as cars can park much closer together and elevators and stairs are no longer required).

Leading smart cities are recognizing that smart parking infrastructure (i.e., sensors and communications networks) can be leveraged to help enable cars of the future to park themselves. AV fleets are also expected to fundamentally change the way cars are used, affecting how often and where future vehicles will be parked. For more information on the smart parking industry, check out Navigant Research’s recently released report, Smart Parking Systems.

 

Blog Articles

Most Recent

By Date

Tags

Clean Transportation, Digital Utility Strategies, Electric Vehicles, Energy Technologies, Finance & Investing, Policy & Regulation, Renewable Energy, Smart Energy Practice, Smart Energy Program, Transportation Efficiencies

By Author


{"userID":"","pageName":"Automated Vehicles","path":"\/tag\/automated-vehicles","date":"4\/23\/2017"}