The amount of data that self-driving cars are predicted to use is massive, really massive. A study by Intel, discussed by Intel CEO Brian Krzanich at Automobility LA, suggests that just one autonomous vehicle will generate around 4,000 GB (around 4 terabytes) of data every day. And that’s just assuming one hour of driving! For more heavily used vehicles, it’s expected that autonomous vehicles will generate and consume around 40,000 GB (around 40 terabytes) of data for every eight hours of driving time.
There’s a number of reasons for this, chief amongst which is the amount of sensors in the vehicle and the very high resolution needed for autonomous driving to be safe. The cameras alone, leaving the radar and LiDAR out of it, are expected to generate data at a rate between 20 Mbps and 40 Mbps. Further, there are plans to implement live map updates which will be far more detailed than existing satellite navigation style maps. Whereas a satellite navigation map might tell you within the nearest 5 metres where a junction is, live maps for autonomous vehicles are planned to be accurate to the nearest inch and include things like the placement of traffic cones or debris in the road. Interestingly, there has also been talk of crowd-sourced driving data, where autonomous vehicles share information in a peer-to-peer network. If an autonomous vehicle screeches to a halt due to a cat running into the road, a separate autonomous vehicle approaching from a blind corner could be told within a fraction of a second of the potential obstacle in its path.
Wireless communication technologies
However for this to be realised, all of that data needs to be moved around a network, and a wireless network at that which induces a slew of other issues. Current cellular technology, 4G, has a typical download speed in the range of 8 – 10 Mbit/s, and a typical upload speed of 5 -6 Mbit/s. You can see then, that the current cellular network is wholly unsuited for the level of data which autonomous vehicles will create. And without the ability to quickly share large amounts of data, quite a few of the advantages of autonomous vehicles are lost.
Enter 5G, the latest in wireless communication technologies. It is expected, once the technology goes mainstream, that 5G’s data communication speed will be in excess of 2Gbit/s. This puts it in a very comfortable position to handle the sort of data loads that autonomous vehicles will be handling. Further, 5G takes a multi-band approach, which should help spread capacity and handle different client devices appropriately. The technology uses three sections of the frequency spectrum: millimetre wave, mid-band, and low-band. Millimetre wave has by the far the fastest speed, however its penetration depth into buildings and other structure is very poor. In contrast, the mid-band and lower-band components trade speed for an increase in accessibility.
This synergy sits well with autonomous vehicles, which are unlikely to spend much time indoors but will require lots of bandwidth! Moreover, the data requirements of the types of devices which can only use the mid- and lower-band components are expected to be much less. Current cellular technologies comfortably support most smart phone devices and IoT implementations, and so restricting their use to the mid- and lower-band components is not expected to cause any degradation in performance.
Which comes first: 5G or autonomous vehicle adoption?
Adoption of 5G has been sporadic, with many countries appearing to wait for the technology to become more established. That said, the timeline for the adoption of autonomous vehicles looks to be picking up pace. Market researchers estimated the global autonomous vehicle market to be worth around 54 billion USD in 2019, but to have grown by a factor of 10 to 556 billion USD by 2026. It seems then that a chicken-and-egg situation may develop, where the integration of autonomous vehicles into the market is delayed by the adoption of 5G; whilst, simultaneously, there is limited appetite to rush the adoption of 5G along whilst there is no significant market for it.
That said, trends within the IP sphere show strong development of both technologies. At the EPO, 5G wireless networks were identified as a key growth area in terms of filings. Indeed digital communication was the technical field with the most patent applications in 2019, with transport coming in fifth. In another study, the EPO found that the number of applications directed to self-driving vehicles had quadrupled in the years 2011 – 2017.
It seems then that 5G is here to stay as a key enabling technology for autonomous vehicles.
