Nokia has completed the sale of its Here mapping business to a group of German car manufacturers, in a move which ends the dream of creating its own web services platform around the location software, but which may see that platform taking an important role in shaping the future of connected and autonomous car standards.
The Finnish firm paid $8.1bn for Navteq, whose technology became the foundation for Nokia Maps and the cloud-based Here services, and got just $2.8bn back from selling it on eight years later, to a group made up of Audi, BMW and Daimler. The deal, concluded in August, valued the business at €2.8bn, subject to certain adjustments. Following those, Nokia’s net proceeds were €2.55bn ($2.8bn).
This may prove to be a bargain price for the car giants, which have the potential to use the software not just to power their own location services, but to give them a platform they can control in future, reducing their reliance on Google or other Silicon Valley players. As the connected car becomes a major growth area for mobile connectivity, and as much of the value of a vehicle shifts from the car itself into software, there is a tussle between web/IT companies and the car industry itself, to drive the direction of the platforms.
Here will be a useful weapon for its new owners, which want to open up the technology to other auto companies, in order to create de facto standards which would limit the influence of Google and others. In particular, there is intense focus on mapping, navigation and other services for future driverless vehicles. The former Nokia unit said that this is one of its most important areas of development. It currently builds its maps from more than 80,000 sources, including information gathered by its 3D mapping cars and from connected devices and vehicles. This will be broadened to include ‘dynamic’ data, updated in real time from urban connected gadgets to produce high definition maps which are sufficiently accurate to support autonomous driving.
The mapping firm wants to put itself at the heart of nascent efforts to develop standards for “cooperative intelligent transportation systems”, based on its maps and tools, and able to power whole urban transport and road safety platforms, including driverless vehicles, rather than just individual smart cars. This would be a powerful asset for the car industry to control, rather than the Silicon Valley players or the mobile operators, which may have seen intelligent transport as something they would naturally control.
Sean Fernback, president of Here, said in a statement: “The map is evolving into a live representation of the world, giving us a second-by-second view of our cities and road networks. Now we have the backing of three automotive companies which share our view that this map will be life-changing for people: it will power location services that improve mobility for people and enterprises, make driving safer and more enjoyable, and reduce emissions.”
He added that the new owners recognized the need for a “clear separation” from the operations of Here, in order to keep it open for use by other auto firms, by other existing customers such as Microsoft and Amazon, and for additional investors.
Of course, there are many other aspects of the autonomous vehicle which will need developing and standardizing before this becomes a common feature of our cities. And a project at China’s Nankai University is already looking another step beyond the real time connectivity and artificial intelligence engines which will power the driverless car – creating a prototype of a vehicle controlled by the driver’s brain signals.
According to the Reuters news agency, the prototype can stop, move forward and backwards, lock and unlock, in response to brain signals, transmitted using headgear with 16 electroencephalogram (EEG) sensors, which detect brain activity, embedded. A computer analyzes the signals and translates them into the automotive commands. The vehicle itself was provided by local carmaker, Great Wall Motors, which has initially focused this area of R&D on helping physically handicapped people to drive. The team stressed that the work was in its early stages and there were no foreseeable plans to commercialize it.
It does, however, highlight the close links between future connected car evolutions and the technologies of virtual reality and artificial intelligence. These two are increasingly moving out of the realms of gaming and sci-fi and being harnessed, in tandem, to create new and highly context-aware user experiences for all kinds of devices, including cars.
The big web players are investing in VR headgear and huge AI back end platforms, hoping to be able to define and own the next generations of online experience and applications. Facebook acquired Oculus Drift and, along with Baidu, Microsoft, Google and others, is making huge investments in AI. Microsoft, too, has a VR headset platform, HoloLens, and it recently announced a partnership with Volvo Cars of Sweden, which highlights how this technology is being taken up by the auto sector.
The two companies will co-develop autonomous vehicle technology and offer HoloLens along with connected cars and the auto giant’s first self-driving vehicle, the Volvo Concept 26. For now, Volvo will use the Microsoft headgear in the showroom to help demonstrate different features and color schemes to customers in a realistic way – but far more disruptive applications might be envisaged for the future, if virtual and augmented reality can be introduced safely into the driver’s seat itself, linking to the kind of AI-driven platforms that Here is dreaming up in the cloud.
Volvo Cars’ retail marketing director Nina Larsen said: “We imagine a future where people will be able to better understand how our safety innovations can help avoid accidents, how our advanced powertrains are optimized to support people in different driving conditions, and how our new connected services can make life easier and save time.”
And while Google is the leading proponent of driverless cars on the Silicon Valley side, it is being chased by China’s Baidu, another huge investor in AI-driven platforms for future vehicular use. The Chinese search giant said this week that it has completed tests, with a prototype car (a modified BMW 3 Series), on a 30-kilometer route with varied roads and environmental conditions. While early autonomous vehicle demonstrations have run in straight lines, developers like the Baidu team are now working on complex driving actions such as U-turns, left turns, lane changes and on-ramps. The route included highways with the car’s speed peaking at 100 kilometers per hour.
The prototype cars have a small server in their trunks with a graphical processing unit to handle AI workloads (the power consumption of that unit is one of the key challenges, says Baidu). The software is called Baidu AutoBrain.
CEO Robin Li said in May that Baidu might introduce an autonomous car this year, putting it ahead of the 2020 date set by Google and carmakers like Nissan. Electric car maker Tesla recently rolled out some self-driving features in its Model S cars.
Baidu’s chief scientist, Andrew Ng, said in an interview that he hopes to develop vehicles that are “fully autonomous over a limited number of routes”, adding: “If we can have a relatively controlled environment I think we’re in striking distance of putting cars on the road.”
The talk about Baidu’s in-car server highlights the opportunity of these complex systems for component makers as well as AI software experts. While many Internet of Things applications involve very low cost, low margin chips, autonomous vehicles will require expensive elements in which vendors will battle to stake a claim. Samsung Electronics has appointed an EVP, Park Jong Hwan, specifically for that purpose, leading a newly concentrated effort in components for self-driving cars and in-vehicle infotainment systems.
Samsung needs to hit back against LG Electronics, which currently has a lead in this area. Park will coordinate efforts with other group affiliates such as battery maker Samsung SDI and software services provider Samsung SDS. Last month, the firm announced a partnership with Volkswagen’s Audi subsidiary to provide memory chips for that carmaker’s in-car systems.
“Samsung just can’t leave this huge market untapped,” Greg Roh, an analyst at HMC Investment Securities, told Bloomberg. He said car electronics, including advanced semiconductors, car audio and sensors, account for about 30% of total smart car revenues. The global automotive semiconductor market is expected to grow to about $32.7bn in 2016, Samsung said, citing Gartner forecasts. The connected and driverless car evolution opens the opportunity for vendors to import technologies from other markets such as smartphones, and to encroach on the powerful auto industry presence of embedded chips from firms like Freescale/NXP and Infineon.
LG is supplying 11 components for General Motors’ new Bolt electric car, including the electric-drive motor, charger and batteries, the result of a four-year old joint engineering pact. LG reported KRW478.6bn ($406m) in revenue from vehicle components for the September quarter.