Futuro della mobilità: la parola a Ericsson

Futuro della mobilità: la parola a Ericsson
 

Futuro della mobilità: la parola a Ericsson

Introduction

Autonomous Driving promises to increase road safety and to ensure mobility for those unable to drive, or to enable industrial use cases in logistic or enterprise areas like port or warehouses. Substantial cost benefits will also drive a cultural shift of car ownership, as so called “robotaxis” become a viable alternative to private cars. UBS, an investment bank, forecasts [nota 1] that autonomous driving, electrification and competition will make owning a car almost twice expensive, compared to an on-demand self-driving service; according to the same report, self-driving shared vehicles could even become cheaper than public transport in Europe. So many Autonomous Driving projects are on the way at OEMs, internet giants (Google, Apple), or mobility providers (Uber, Lyft), that one could argue that self-driving cars are around the corner, but the industry is still facing some technology, regulatory and cultural issues. Video and Radar have been combined with deep-learning inferencing algorithms, to scan vehicle proximity and take autonomous decisions about accelerating, braking and steering, and these advances have gone a long way to climb up the levels of autonomy, defined by the Society of Automotive Engineers (SAE J 3016-2018), which range from Level 1 (basic services like Cruise Control or Lane Keeping Assistance) up to Level 5, which makes human intervention unnecessary under all roadway and weather circumstances (see picture 1).
Higher levels of autonomy prove hard, especially when roads are not as predictable as straight highways, with clearly delimited lanes.  Safety problems loom, when drivers must revert to manual mode but may lack the required readiness, if busy reading or texting. As a matter of fact, Autonomous Driving will come first where road conditions are predictable, and automation brings a clear business case, such as for enterprise vehicles that operate in specific areas like campuses, ports, logistics terminals, mines or bus lanes.

 

Autonomous Vehicles must be connected

Despite progress of on-board technologies, the large scale adoption of Autonomous Vehicle will ultimately depend on: 1) regulation and control of conditions that allow Autonomous Driving; 2) availability of high-definition maps that enable precise localization; 3) customer acceptance, based on intuitive Human-Machine Interactions; 4) collection of driving data, to feed the improvement of intelligent driving algorithms, when the system uploads sensor data in case of hand-over to manual mode. Vehicle Connectivity can help to address all four factors listed above, with various Vehicle-to-Everything (V2X) use cases, that include Vehicle-to-Vehicle interactions (V2V), communication with Road Side Infrastructure (V2I) - traffic lights or meteo stations etc., or cloud-based services via Wide Area Networks (V2N). Low latency V2X is a key enabler of advanced active safety, complementing on board sensors with non-line-of-sight visibility. V2N will be crucial to control the hand-over to the human driver, and for the improvement of robotic driving. Cellular V2X (C-V2X) specification, part of 3GPP Release 14, was completed in 2017, and it will evolve into 5G with Release 16, expected in 2019.
It is remarkable that Autonomous Driving comes along in sync with 5G deployments, that offer unprecedented performance levels. Flexible allocation of cloud workloads and reduced latency will enable mission critical functions, while enhanced data rates will fit to automotive needs, as experts expect Autonomous Vehicles to upload more than 1TB of video, sensor and telematics data per month by 2025 [note 2].

 

5G-Automotive Association

The automotive industry has been considering the adoption of 5G, and the 5G Automotive Association (5GAA) has gathered industry players around this topic. 5GAA was initially established in 2016, by few founding members that include Audi, BMW, Daimler, Ericsson, Huawei, Intel, Nokia, Qualcomm. Today more than 90 participants include prominent telecommunication providers such as Orange, SoftBank, Telefonica, Deutsche Telekom and Vodafone. This participation shows that Telco and Automotive leaders share the need to ensure the interoperability of 5G and to shape future use-case ahead of 5G network roll-out.

Ericsson, a founding member of 5GAA, has been directly involved in multiple Autonomous Driving projects. Since 2017 Ericsson collaborates with Zenuity, a joint venture between Autoliv and Volvo, to develop an end-to-end platform for Autonomous Driving and Advanced Driver Assistance. Other relevant Ericsson engagements are described in the following paragraphs.

 

Autonomous Driving Aware Traffic Control in Gothenburg

Autonomous Driving Aware Traffic Control is joint public/private program, established in 2017 and co-financed by Vinnova, the Swedish National Innovation Agency. The partnership includes Volvo, Ericsson, Carmenta, a supplier of software for mission-critical systems, Trafikverket, the National Transport Authority, and the City of Gothenburg. The main deliverable has been a demonstration of the information flows that are exchanged by Autonomous Driving vehicles, the Central Traffic Control system and Road Authorities, to improve safety of Autonomous Driving.
The Volvo vehicles aim at Level 4 automation (= secondary tasks are allowed) and when they cannot manage the situation, they go to a safe stop or make a controlled handover to the driver. Autonomous Driving is allowed on a carefully mapped set of road segments, but not under severe weather or traffic conditions.
In the future OEM´s will share the same need for a common authoritative and certified source of traffic and weather data, to base their decision about allowing or revoking Autonomous Driving. The Central Traffic Control system can serve multiple OEM’s, by aggregating and sharing the data of interest, based on the DATEX II standard (CEN/TC 278). The Autonomous Driving Aware program proved that real-time traffic awareness is a cornerstone for safe and efficient traffic in self-driving mode.

 

AD Aware system architecture

C-V2X demonstration in Europe - Convex

Established in 2017, the Connected Vehicle to Everything of Tomorrow (ConVeX) program has the goal to operate a testbed for Cellular V2X, and to validate its performance. The associated consortium consists of Qualcomm, Audi, Ericsson, Swarco, the University of Kaiserslautern, and the German Federal Ministry of Transportation and Digital Infrastructure (BMVI). In 2018 the program has implemented a live demonstration of direct communication between Motorcycles and Vehicles, equipped with C-V2X technology based on Qualcomm chipset solution. The test scenarios include “Intersection Collision Warning”, where a car pulls out from a junction and avoids hitting a motorcycle rider, as well as “Across Traffic Turn Collision Risk Warning”, in which a vehicle avoids a left turn collision with a motorcycle. V2N communication was run on Ericsson’s 4G and 5G cellular test network, to support telematics, infotainment and safety.

 

Remote operation of vehicles with 5G

In order to make Autonomous Driving feasible in Public Transport and Road Freight in the near future, either a driver needs to be on board - ready to take over - or the vehicles need to be connected and controlled at a distance. The presence of remote operators scanning screens and ready to intervene, can contribute to public acceptance. Remote operation requires broad coverage, data throughput and low latency to enable exchange of HD video feeds and commands between a remote operations center and a vehicle, therefore Scania has implemented a cellular test network, in collaboration with Ericsson, to measure total system response time, and to assess the provisioning of prioritized network services, that are needed to protect quality of service for mission critical applications. Network delay (RTT) with LTE radio mostly stayed under 50 ms, during the study. Subsequent tests with 5G radio access, achieved network RTT to under 4 ms. The testbed includes an evolved 5G core and allows automated service ordering and provisioning, allowing the set-up and take-down of prioritized network resources, if remote operation is required, with specific QoS, such as designated latency levels and guaranteed throughput.

 

Conclusions

Autonomous Driving will be a transformative technology, as much as the smartphone was in the last decade. 5G will bring several benefits to Autonomous Driving and remote-control systems, including core network slicing that will enable priority service provisioning. The convergence of autonomous driving and 5G offers to Mobile Operators the opportunity to work with automotive players and Transport Authorities to develop use-cases, ahead of 5G network roll-out. 5GAA represent a unique cross-industry board where Mobile Operators can influence the global standardization of automotive technologies.

 

Note

  1. UBS – Smart Mobility Report - 19 October 2017
  2. Automotive World ldt 2018 – Special Report on 5G and the autonomous vehicle
 

comments powered by Disqus