Highly Automated Driving, not just a buzz

By Jan-Maarten de Vries, TomTom

Automated driving is a very hot topic in the industry and we have been talking about driverless cars for some years now. With it comes the biggest paradigm shift in the automotive industry since we moved from powered by horses to “horsepower” more than 100 years ago. The driver will become a (co-)passenger and the driving task will be transferred to machines.

Automated driving will disrupt the automotive industry as we know it in ways we have not seen before and cannot even completely conceive today. The automated driving arena will soon become a complex ecosystem of many different cross-industry players and technologies. End-users will benefit from increased safety, comfort and fuel efficiency while the impact on society will be equally as big; mobility will become accessible to all and traffic and parking problems will be greatly reduced.

At the same time there are still some barriers to overcome, among others around safety and liability – do computers indeed drive safer than humans and who is responsible in case an accident does occur? Legislation will need to be adjusted and what about user adoption – people can save time but where is the fun?

Today, cars are already becoming more and more automated. Advanced Driving Assistance Systems (ADAS) features are increasingly being introduced in new cars. While some ADAS applications rely on sensors only, many are now taken to the next level by using very detailed, up-to-date and richly attributed map and/or highly accurate traffic information. Advanced intelligent cruise control, upcoming curve alerts and jam tail warnings are just a few examples.

As the next steps are taken towards Highly Automated Driving, many more technology components come into play. Multiple redundancies between these systems are needed to create solid and safe HAD systems.

Most people in the industry recognize that a highly accurate, detailed and up-to-date map and accurate travel services are key products to bring HAD to reality. They allow the car to see beyond the reach of its sensors.

Our role in automated driving is to help our partners (car makers and system vendors) developing vehicles capable of answering to these simple yet complicated questions: (i) where exactly is the vehicle located? An HAD map provides highly detailed road context and allows precise positioning of the vehicle with decimetre precision; (ii) What lies ahead? Connected services such as traffic and advanced weather services provide live road status beyond the reach of vehicle sensors; (iii) Where can the car go? Navigation software and eHorizon calculate the most probable paths and guide the vehicle to its destination. Finally, based on driver speed profiles and local point speeds, comfortable speeds need to be identified so the driver can reach their destination in the driving style he or she likes. A big challenge for map makers will be to keep the HAD maps up-to-date. Automated vehicles will heavily rely on fresh maps as roads change all the time. It is therefore crucial to add near real-time car sensor data to the map making process and to have access to a mapping platform that can make releases on a continuous basis. Years of experience in processing data helps a great deal in making this happen.

Many different parties will be involved in making HAD a reality. Next to car makers and Tier1s who can normally push for innovations, there is a big dependency on governments worldwide. Allowing self-driving cars on the same road as normal cars is a big step. We believe that eventually, as technology matures, everyone will agree that automated driving technology will be safer than humans, driving tasks will progressively be transferred to machines and adoption of HAD systems will increase. In the meantime every test drive is bringing us closer to the future.

AUTONOMOUS CARS AND THE RISE OF NEW MOBILITY BUSINESS MODELS

‘Status of the autonomous car based on MWC feedback – including a coverage of V2I/V2V connectivity’

By Pietro Boggia, Frost & Sullivan

The concept of autonomous driving vehicles has been thoroughly discussed in terms of advancements of enabling technologies involved. The terms of Autonomous Cars and Driverless Cars are often confused: complete driverless vehicles in the true sense are still a long way off, but the penetration of the enabling technologies is increasing constantly in the market.

Advanced Driver Assistance technologies (ADAS), such as lane departure warning, automated park assist, blind spot detection, and forward collision warning are already being considered a standard in cars. In addition, connectivity brings on the scene advanced communication technologies between Vehicles (V2V) and between Vehicles and Infrastructure (V2I), to assist in overtaking, emergency breaking, turning at congested intersections, weather and road geometry information among others. All these advanced driver assistance features are paving the way for the fully automated cars of the future, where the driver will be present in the car to take on the controls only when required. Don’t expect Driverless Cars to experience a breakthrough market entry, as infrastructure, policy, insurance and ultimately the end-user won’t be ready for that for the next few years. Think more about an incremental wave of innovation from ADAS technologies, to semi-autonomous vehicles with increasing technology penetration for progressive driver substitution.

OEM Activity

One of the OEM that is leading the way on this field of Autonomous Cars is certainly Tesla, which recently announced the launch of an over-the-air software upgrade this summer, enabling all existing vehicles to run in “Autopilot” mode towards the nearest charging station, during freeway driving mode but also assisting during parking mode. What is really surprising about this announcement is that this is a car we can buy today, which will be given the ability to drive itself in a few months via the same setup, very similar to what happens when we receive updates for our iPhone.

In the meantime Google is starting to open up about its driverless car project, which has already put hundreds of vehicles on roads in California and Nevada. The Google approach is much more drastic as the company is working on building a fully self-driving car, without even being fitted with a steering wheel. The ultimate goals behind this revolutionary idea are the increased safety for passengers and the savings on time and costs spent on traffic. Google is simulating thousands of real-life situations to make this car a reality. In the future, V2V communication will also enable vehicles “to learn” in real time from each other in different challenging real-life situations not predictable by testing or simulation.

In the UK, self-driving prototypes will move from off-road to on-road activities this summer, thanks to a Department of Transport ruling. We’ve already seen the self-driving Audi at the Consumer Electronics Show in Las Vegas this past January; that too has been tested in California. BMW and Mercedes are also on the case, while chipmaker NVidia has dedicated its most powerful chip ever to the forthcoming slew of driverless cars.

Feedback from MWC

In fact, in shows such as the CES in Las Vegas or the most recent Mobile World Congress in Barcelona, where some OEMs announced their (not so) future plans and news about autonomous cars, are quickly replacing the most common auto shows such as Detroit or Frankfurt on the headlines of new technology announcements.

This includes Renault-Nissan, which CEO Carlos Ghosn announced that the company will launch autonomous cars in 2016: the autonomous vehicles differ from true driverless vehicles, such as the ones Google has been testing in California, as humans retain an element of control. Ghosn said they will limit accidents and make driving more pleasant. Volvo is going to test driverless cars in Sweden: speaking at Mobile World Congress, Klas Bendrik, CIO of Volvo, said the company is developing a driverless car, which would be tested by “normal people on normal roads.” In February, Volvo revealed details of its “Drive Me” program, which will see the company work with transport authorities and legislators to get driverless cars on the roads as soon as 2017. Ford pointed out that while technology will be ready soon policy and insurance will not be.

Key Questions

Some key questions are coming along with the driverless vehicle concept:

• Why the need for an autonomous car? Key reasons are:
  • Saving lives
  • Saving time
  • Saving money
• Will autonomous cars be the same as cars we see on the roads today?
  • Not necessarily, as Autonomous drive will open up the possibility for OEMs to explore new product concepts designed around new business models, extending the range to different market niches including electric micro vehicles aimed for specific customer segments, eg. elder people.
• Will Driverless cars come in increments or as a disruptive innovation?

o   They will likely be introduced as an incremental penetration of autonomous technologies, driven by OEMs as an industry push to continuously differentiate their product offering, until the regulatory frameworks and customer acceptance will be ready to generate a broader demand in the market. In this context, Autonomous highway driving is due to become a mainstream by 2020, while driverless trucks will also be introduced for platooning towards 2022.

What seems to be a worrying trend for automakers is that self-driving cars will bring with them a new concept on how we use, and eventually buy, cars, as self-driving vehicles are the ideal platform to build new business models on it, related to the sharing economy and the concept of mobility on-demand. This may eventually result in a threat for mass-makers as with fewer cars on the road, less differentiation among autonomous vehicles based on vehicle performances, selective purchasing preference will likely favour premium brands where a larger amount of differentiating new technologies will be available compared to the offering from low-end makers.

Collaboration will be key

• “There are so many unknowns and we cannot solve all the problems ourselves, so there are a lot of opportunities for start-ups and competitors that invest and develop new technology”. Says Ghosn. The rise of Auto IT and related technologies, such as Connectivity, Autonomous Cars and V2X communication (also called Cooperative Systems) pose a serious technology challenge to OEMs, as they have a clear lack of internal skills and require suppliers to cover this gap. But unlikely is the continuation of the traditional Automotive business model, where OEMs traditionally retain the branding and the overall added value on the product; new technologies are pulling the role of OEM away from the centre stage of value creation in favour of Solutions Providers, Mobility Integrators and in general Technology Providers.
• As clear example of this trend, Fiat Chrysler Automobiles CEO Sergio Marchionne said he is open to an alliance with Apple or Google as he seeks a partner to merge with.

Driverless cars only have a problem: we love driving. Will OEMs start to offer driving simulators to their customers to keep them loyal to the brand and products? Will the automotive industry use gamification to remind their customers the pleasure of a good ride?

I personally believe that at a certain point traditional OEMs will lobby against the totally driverless vehicle, which I would not call a “car” as we usually define it, as they won’t be ready to drastically change their business models and their customers’ acceptance towards a new era of driverless ride.