5G & Autonomous Cars: Flashy Promise Meets Complicated Reality
Autonomous driving is a big part of the story being told about 5G. Almost every vision of 5G includes it as one of the most compelling applications for the next-generation standard. But, as with so much about 5G, the reality is more complicated than plugging in a new network and watching self-driving come to life.
The robocar applications promised with 5G sound great: 5G's ultra-reliable, low-latency communication (URLLC) -- coming in 3GPP Release 16 next year -- might give self-driving vehicles access to cloud-based intelligence at the speed of a split-second driving decision. Enhance mobile broadband (eMBB) would allow them to download software updates on the road. By constantly sharing real-time location and driving intentions through direct vehicle-to-everything links (V2X), the cars could stay out of each others' way.
But even if those applications become real, it won't happen without a lot of work on the back end. For one thing, having useful 5G links among autonomous vehicles will depend on mass adoption of connected cars to enable a network effect. Even more daunting, building 5G operator networks with the low latency, high throughput and solid reliability that AVs need will require changes at multiple layers.
The relationship between 5G and AVs will be less like a quickie wedding at a drive-through chapel and more like a long family road trip. Between still-emerging technologies and difficulties in deployment, the large-scale use of 5G for vehicle automation probably won't happen for another five years or more.
To begin with, there's no industry consensus about either the basic requirements of full self-driving or how it will use cellular networks. Tesla and Waymo, for example, are mostly relying on powerful onboard hardware and software. Established automakers such as Ford and Volkswagen are exploring a mix of in-car computing and connected features, including network-based services and V2X, a set of applications using direct wireless links to nearby vehicles and roadway infrastructure. (See Sidebar: C-V2X)
Even the best network won't be enough by itself: AVs will need to be able to drive where 5G coverage is spotty, so they will also have enough onboard capability to use sensors and V2X to operate safely, said Jovan Zagajac, Ford Motor's manager, Connected Vehicle Platform and Product. But where available, cell networks could play important roles in applications such as software updates and remote operation, he said.
By turning to network-based services for things like real-time data for decision-making, automakers will be able to reduce the amount of computing power and software intelligence they build into their vehicles for making driving decisions, said Martin Beltrop, head of Nokia's mobile networks automotive business. "We would benefit from the information we could collect in the cloud to simplify the decision," he said. But he added, "the final decision to go or stop will be done in the car."
Over time, 5G mobile operator networks could host an array of services supporting large-scale use of AVs. The key features to make these services possible may include network slicing, edge computing and even new business relationships.
Learning from the present
An early example of network-supported self-driving, currently using 4G, illustrates why 5G may play a key role for AVs in the future.
Swedish mobile operator Telia is piloting a service that supports driverless trucks hauling goods between warehouses. It has been operating in a commercial pilot at a logistics facility in Jönköping, Sweden, according to Ericsson, which supplies infrastructure for it.
The driverless trucks, developed by Swedish startup Einride, operate on pre-defined routes as part of an automated logistics system. The trucks can make most of the trip on their own with inputs from onboard sensors, but there are some situations they can't navigate, said Claes Herlitz, vice president and head of Global Automotive Services at Ericsson. When that happens, a human driver in a remote operations center takes over. Remote human driving, or teleoperation, will be a mandatory feature of AVs in many jurisdictions, Herlitz and others said.
In the Einride case, teleoperation requires both high upstream bandwidth to stream live video from trucks and low latency to ensure driving commands arrive in time. In Jönköping, the service is running on an advanced 4G radio network with a 5G core. The network has end-to-end latency of about 15ms and supports the service well, Herlitz said. However, with LTE, it has only been possible to stream high-definition video from four trucks per cell.
5G radio networks will allow this type of service to scale up to more vehicles, thanks to new spectrum bands, beam-forming, MIMO techniques and improved user-equipment capabilities that will provide higher capacity, Ericssson says. 5G core capabilities including URLLC and network slicing will help to ensure low latency.
If connected and autonomous vehicles ever dominate the roads, if may be possible to make driving even safer by managing them as a system. 5G WANs may deliver services that enable this mass automation.
Here's how one example of this might play out: Picture a major city intersection where as many as 1,000 cars, bicycles, scooters, pedestrians and other objects share the space at a given time. Watching over the crossroads are 20 cameras to monitor any participants that aren't equipped to communicate their own location, speed and intent.
An automated traffic management system collects the data transmitted by these objects and sends all relevant information to AVs, each of which uses those inputs in conjunction with its own sensors to decide whether to stop, go or turn. The data needs to be received, processed and sent out in near real time because the traffic situation is constantly changing.
Such a system couldn't work without a 5G network, according to Nokia's Beltrop and other automotive and mobile executives. The number of connected objects, the volume of data and the tiny margin for delay would require automakers and mobile operators to leverage new capabilities that are just beginning to emerge.
Some see 5G-enabled AVs getting even closer to networked robotic control. Road operators, collaborating with mobile carriers, could set up traffic management systems that effectively automate activities such as merging, said Maxime Flament, CTO of the 5G Automotive Association (5GAA), a cross-industry group backing automotive 5G.
Connected vehicles would communicate their location, speed and direction to the management software, which would create a model of the overall traffic situation. (Sensors in the roadway and in nearby cars would monitor unconnected, human-driven vehicles.) The system would create a virtual contract with each connected AV, in which the vehicle would commit to taking an action such as stopping, accelerating or turning. Based on those commitments, another AV coming into the roadway could safely merge into the flow of traffic, Flament said.
Next page: Network reliability and a new role for carriers
Network Reliability & A New Role for Carriers
The latency breakthrough
Low latency is the 5G capability attracting the most attention for vehicle automation. While there's no universal agreement on what's needed, a frequently cited target is end-to-end latency of 10ms or less. Depending on how much an AV's onboard computer relies on network-based services for decision-making, the requirement could be as low as 2ms, Nokia's Beltrop said.
That's much tighter than the requirement for voice-over-LTE (VoLTE), one of the main low-latency applications of 4G. VoLTE can work with a round-trip latency of about 100ms, said Cameron Coursey, vice president and CTO of IoT at AT&T. This is where 5G becomes a windfall, because the new specification is designed to provide much lower latency, beginning with RAN latency as low as 1ms, compared with tens of milliseconds on average with 4G.
Another important piece of 5G for low latency is its flexible network architecture, which will let operators place computing resources toward the edge of the network to avoid long round trips to distant cloud data centers. More on that in a later article.
Big driving data
Other AV applications may need 5G for sheer speed. All self-driving vehicles run complex software, especially DNNs (deep neural networks), that need large updates. The cars also continuously collect huge amounts of onboard sensor data, plus information about the outcomes of driving decisions, that automakers and suppliers can collect to improve the self-driving software. These frequent downloads and uploads will benefit from the gigabit-speed wireless connections that 5G carrier networks are designed for.
Software updates may be frequent, and centralized systems may collect and analyze driving data quickly, but these are unlikely to happen in real-time. As a result, operators won't necessarily need to provide that kind of broadband to AVs in motion. Big data transfers could be activated when the vehicle is stationary and in range of a high-speed connection (at a charging station, for example), especially in areas with dense, high-frequency 5G coverage.
Another automation concept calls for AVs to share real-time sensor data so a car can "see-through" the vehicle in front of it, especially a large truck that blocks the view of traffic ahead. Because it involves real-time streaming video, this is likely to require both high-speed broadband and low latency, with assistance from edge computing, Flament said. As a result, this application would probably go over a WAN instead of a direct V2X connection, so roadside networks would need to get faster and more robust, he said.
Making sure the network comes through
Reliability is a major concern for autonomous driving applications since a loss of signal -- or of a driving assistance application -- could affect safety. One significant 5G advance to help ensure performance and availability is network slicing, which will let mobile operators set a specific quality of service for an application by assigning virtualized network resources to it. An AV application could be assigned priority over other network applications due to that safety requirement. This is one thing Ford, for example, is seeking to ensure is supported in its AVs.
If autonomous driving services need guaranteed QOS on 5G networks in the next five years, they will probably use generic network slices like those used by other applications running on shared edge servers, 5GAA's Flament said. This edge computing infrastructure, along with network slicing, will be deployed first in dense urban areas where there are enterprise customers for it. Later, when 5G comes to highways outside those areas, carriers may create specialized network slices for AVs if a road operator requests them.
Network slicing is a start, but making 5G networks reliable enough for self-driving will be a tall order, analyst Philip Marshall of Tolaga Research said. Achieving low average latency isn't enough to support mission-critical applications like driving, Marshall said. What's needed is consistent low latency. That will require much higher network density than some people expect, including redundancy in both cells and computing infrastructure -- the lower the latency needed, the higher the cost, he said. Also, large-scale implementation of network slicing could take several years and hasn't yet begun, Marshall said. To achieve results without building separate infrastructure for priority applications, it will require virtualization of both the core of the network and the RAN. To use network slicing for self-driving applications on the open road, carriers first would have to convert to a cloud RAN architecture over a broad area, and it's way too early for that, he said. "Network slicing will get incubated with applications that don't have wide-area requirements," Marshall said. "To try to implement network slicing over a wide-area 5G network environment is crazy."
New role for carriers
Network uptime will become a bigger issue when 5G networks start supporting autonomous driving, Ericsson's Herlitz said. That may start with early deployments for enterprise customers such as logistics and waste management companies.
Such customers will need service-level agreements (SLAs) better than anything offered now, particularly with regard to fixing outages, because most carrier SLAs today are geared toward consumers, he said. In this case, the mobile operator will need to be integrated into the customer's business, with technicians on site to solve problems that affect network availability immediately. When a customer's core business relies on robotic trucks, 48 hours without service is prohibitively expensive, he said.
For times when AVs travel beyond the reach of all this infrastructure, automotive supplier Continental has demonstrated a system to help the vehicle prepare. "Predictive connectivity" uses historical information about network performance and predictions of a vehicle's route to gauge where the vehicle might run out of coverage, Continental says. Then the car can change to a different network, prioritize the applications in use, or even shift to a failover mode in which it relies on built-in sensors and computing power.
As for whether mobile operators, automakers, application providers or other entities will be held responsible if a network-dependent self-driving application fails, it's too early to know, Herlitz and others said. Even without the network element, liability for traffic incidents involving AVs is already a hot topic that's far from being resolved.
Setting a timeline
Bringing together 5G and self-driving cars will involve multiple daunting technology missions for players in both networking and automotive. 5GAA is working with network vendors and automakers to converge their timelines so neither cars nor networks get stranded waiting, Flament said.
While the first 5G modems should start showing up in new vehicles in 2022, ones that support URLLC and new, high-frequency radios probably won't arrive until 2025, Flament said. Meanwhile, 5G infrastructure that can talk to those modems to enable self-driving isn't due for commercial deployment until about 2025, Flament said. A new generation of V2V that lets vehicles share more data, more reliably, to better support self-driving, may arrive around the same time, he said.
As 5G expands the role of mobile networks from primarily consumer voice and data services to emerging and mission-critical such as automated driving, network demands are growing and infrastructure becoming more complex. While the roadmaps for both 5G and self-driving are still being drawn, there's a clear possibility that 5G will help drive vehicle automation forward.
In the rest of this series, I'll look at more details about two evolving aspects of connected self-driving: edge computing and wireless spectrum.
Next page: Sidebar on C-V2X
C-V2X: Direct Links to Augment Network-based Automation
While some uses of 5G for vehicle automation may place new demands on operators' networks, other features would use direct links among vehicles, roadway infrastructure and pedestrians' devices.
Qualcomm, BMW, Ford Motor and other companies are working to extend cellular vehicle-to-everything (C-V2X) onto 5G. The technology, which is now being tested on 4G radios, is a cellular alternative to V2X systems based on the IEEE 802.11p wireless LAN standard, including Direct Short Range Communication (DSRC).
Both systems let connected cars and other road users communicate even where cellular network coverage is weak or nonexistent. They have several modes of operation designed to make connected human-driven cars -- and later, autonomous vehicles (AVs) -- cooperate for greater safety.
For example, AVs could warn each other about obstacles such as construction zones or stalled cars ahead. Traffic lights could communicate their status to AVs and time their changes to improve traffic patterns. Emergency vehicles could force AVs to pull over as they pass. V2X-equipped smartphones in pedestrians' pockets could notify cars when the walker is crossing the street. Roadside beacons could send traffic and weather information and live, high-definition maps to passing AVs.
Proponents say C-V2X has longer range and higher reliability than 802.11p, plus an evolutionary path to 5G. It may also be able to piggyback on in-car modems provided for entertainment, information and diagnostics over 5G.
Enhancements targeted for Release 16 of the 3GPP's 5G standard would allow for latency of less than 1ms in C-V2X, according to Qualcomm. But the US and European Union are still weighing which technology to mandate or endorse. The wireless LAN-based systems have been in the works for years, though they haven't been widely deployed.
Last week, Federal Communications Commission Chairman Ajit Pai proposed a shift in U.S. policy that could be a big win for C-V2X. Pai asked the FCC to consider reassigning a 75MHz chunk of the 5.9GHz spectrum band that it allocated to DSRC in 1999. Saying DSRC had failed to reach its potential, he proposed handing over the bulk of that spectrum to unlicensed use, especially for Wi-Fi, and dedicating 20MHz of the band to C-V2X. The other 10MHz would be allocated to either C-V2X or DSRC, depending on public input.
C-V2X can achieve consistently low latency regardless of congestion, according to the 5G Automotive Association (5GAA), an industry group that promotes automotive uses of 5G. Qualcomm has said the system can provide latency of 4ms or less depending on the implementation. To meet latency requirements, the system can optimally allocate resources within a spectrum band shared by multiple vehicles using deterministic resource scheduling, Qualcomm says.
While C-V2X has a mode in which the direct links between vehicles can be managed by basestations on a cellular network, it also allows onboard radios to manage it all on their own. With no need to send packets across a carrier's network and onto the cloud, low latency is easier to achieve.