Incoming 5G networks, buoyed by network slicing and artificial intelligence (AI) techniques, will be at the heart of developments in autonomous vehicles, intelligent transport systems, and driver safety, BMW Group has said at a 5G demo with Nokia at its headquarters in Munich.
The German car maker said it is already developing 5G systems with Finnish network vendor Nokia, already, to be in a position to offer 5G functions in its production vehicles as early as 2020. Autonomous vehicles are “only possible” with the speed and reliability afforded by the new 5G “high-speed information highway,” it said.
“5G mobile networks will be more powerful than ever. Data transfer rates will increase many times over while latency will drop to one millisecond. This new quality of connectivity is underpinned by a new, exceptionally close-knit infrastructure of radio masts, new transfer protocols with high spectral efficiency, and a widely spread and powerful cloud infrastructure,” it said in a statement.
“5G opens up new possibilities for vehicle connectivity. The data rates it achieves will take the quality of in-car online entertainment to a whole new level, while connectivity between vehicles and with transport infrastructure can also be stepped up.”
In particular BMW put focus on the application of 5G network slicing and AI in the automotive space. In Munich, it demonstrated how individual 5G slices might be carved up, with one slice used for updating HD maps, a second reserved for sharing time-sensitive data between vehicles, and a third optimised for streaming HD videos.
BMW said 5G will be crucial to the implementation of wireless cellular vehicle-to-everything (C-V2X) technology for two-way communication between road users. It said it will use C-2VX to augment data insights for drivers, beyond details about filling stations and car parks, adding new information services from mid-2019.
Data from cameras and sensors in BMW cars will be re-purposed in dashboard displays to give drivers warning about emergency braking ahead, or police requests for emergency corridors in the traffic. “Data gathered is gleaned from the cameras and other sensors installed in the vehicles, from where it is transmitted in entirely anonymous form via the built-in SIM card.”
For data processing in autonomous driving systems, AI is of “instrumental importance,” it noted. “These systems will be refined to the point where the vehicles equipped with them can safely cope with even the most complex driving situations in urban areas,” it said.
“The development goal for the longer term is fully autonomous driving, when all of a vehicle’s occupants can ride as passengers without any driving tasks to fulfil. In a self-driving car, artificial intelligence is the key to safe and comfortable mobility for all.”
BMW also showcased usage of mixed reality simulations, with borrowed techniques from the consumer electronics and video game industries, in its manufacturing processes, providing a means to visualise designs before production. “This entails far less work than constructing multiple physical prototypes. What’s more, the virtual experience can even be shared by several developers over long distances,” it said.
BMW is also innovating with additive manufacturing, or 3D printing, it said, giving it greater freedom with the design and manufacture of car parts, particularly of custom-made and highly complex components, required in small numbers. The technique is most applicable in pre-development, vehicle validation and vehicle road testing.
It supplies around 140,000 3D-printed prototype parts per year from its Munich plant to its various development departments. The aluminium mounting in the soft-top cover of its i8 Roadster is produced by means of metal powder laser melting, it said, the first time the technique has been used in car making.
Spanish telecoms group Telefónica and vendor Huawei completed what the companies the world’s first proof-of-concept for 5G based V2X networks in their joint innovation lab in Madrid, in Spain, in January.
Korean telecoms operator SK Telecom completed a 26-kilometre trial of a self-driving vehicle in September last year, in real world traffic conditions at a speed of up to 80 kilometers per hour. General Motors ran its own experiments with data communication between the traffic signals and the vehicles in June last year.
