Local Intelligence’s Role in Fully Automated Mobility
The path towards automated, autonomous mobility is complex to the extent that there is no single path to achieve it.
However, this transition it is either viewed as an automaker initiative or as a technology disruptor. Location intelligence plays its part in this process. It entails the collation, analysis and distribution of geospatial data to provide a digital representation of the environment to increase situational awareness of automated and autonomous vehicles.
Evolutionary path
Jørgen Behrens, senior vice-president and chief product officer at HERE Technologies adds that it involves both an evolutionary and revolutionary path. He explains: “The evolutionary path means adding new automated driving capabilities to passenger and commercial vehicles, to increase the safety and convenience by taking over routine driving tasks from the driver. However, there is still the need of a driver supervising and taking over, which is why it is most often referred as conditional automation. The advantage of this path is that can be deployed at scale, and it is relatively affordable to own.”
Revolutionary path
“The revolutionary path aims to replace the driver all together and it pursued as an alternative to human chauffeurs or drivers in ride hailing services and geofenced logistics. There are multiple societal benefits for highly or fully autonomous mobility, but due to the costs and complexity of operating without human supervision it will probably take some time until this will become a mobility solution available at scale.”
Mark Cracknell, head of connected and automated mobility at Zenzic, reveals that his company is working on The UK Connected and Automated Mobility Roadmap to 2030, which establishes the path and vision for the connected and automated future of mobility in the UK. In terms of evolutionary paths, or even revolutionary paths, he says the report contains four overarching themes.
He said: “The first being society and people; ensuring public acceptance, legislation and insurance and licensing and use, including vehicle approvals. The second one is vehicles: their software, machine learning, hardware, and sensors. The third is infrastructure, and underneath that is communications, the emerging road infrastructure, and the digital infrastructure. The fourth and final path is around services.”
In essence the paths are about how the vehicles are used, and type of mode of transport employed as part of an integrated pathway to fully automated mobility. For example, this could involve passenger transport with, for example, autonomous taxi or bus services, or freight and logistics such as an automated delivery van.
Communication
A key backbone to the integration of these different modes of transport is communication, and the ability to provide local intelligence to each vehicle and throughout the system. Speaking about why there is a need for local intelligence, Cracknell adds: “There is an opportunity with connectivity and autonomous to provide transport equity for the disabled and elderly, or for people who feel the cost of transport is high. They can provide a lower cost transport service in a broader area.”
He explains that there are two ways autonomous vehicles can navigate the roads and the environment around them. The first is in real-time, reacting according to what the vehicles see, and the second is local intelligence, which involves using high-definition mapping to allow the vehicles to know their environment in advance. Cracknell says this means that there is a base level understanding of the static environment before the vehicles have reached it. “This can permit a higher level of confidence and safety in your vehicle,” he says.
Kersten Heineke, partner in the Frankfurt office and leading the McKinsey Center for Future Mobility at McKinsey & Company, concurs that there are many elements to local intelligence. Beyond the HD maps, there is also an opportunity to provide location-based services that accumulate data from the environment sensors, or in-vehicle sensors. They can then provide insights from them to other vehicles – including road hazard warnings. Remote control centers offer an additional layer of supervision for autonomous vehicles.
While they will not continuously monitor cars, they will offer remote guidance in case the AVs request remote supervision to approve, for example, certain maneuvers. Then there is smart infrastructure, which he says is connected to the vehicle via Wi-Fi or via the cloud to, for example, operate and control smart traffic lights.
He adds: “Most of these elements are not required for autonomous driving but can accelerate the introduction of autonomous driving, by reducing the number of difficult edge cases. If cars need to rely on local intelligence, the connectivity needs to be guaranteed (which might require redundant solutions). The remote-control center might be required by regulators.”
Planning and localization
Behrens says location intelligence is used by a wide array of vehicles and automated driving capabilities. He believes there is a clear case for automated and autonomous vehicles to use it to enable and enhance:
- Planning – to define the most efficient route and path between two points
- Localization – to provide information on the roads, lanes and road furniture (poles, dividers, bridges, etc.) that vehicles can leverage to increase positioning accuracy.
He also suggests that location intelligence can be used to complement sensors’ inputs. For example, he reveals that from Mid-2022 new vehicles in Europe are required to have an intelligent speed limiter that warns or prevents the driver to go unintentionally over the know speed limits. Yet the implementation of this technology is not so simple.
In fact, he believes it’s actually quite complex. He explained: “The complexity is that a large part of the applicable speed limits are not posted as a numerical sign or they may be temporarily blocked or damaged. So, in order to meet this requirement there is a need to operate with location intelligence, in the shape of speed limit information coming from a map or a connected service.”
Key strategies
Cracknell suggests, amongst others, they are the following:
- Establish the regulations to enable local intelligence and automated mobility. In the UK, for example, there is much work going in CCAV, a government department, who are leading a program to define the regulations and legislation.
- Maximize the opportunities for data sharing as there are lots of opportunities to share data, connect systems and cities together but that only works if people are motivated to share that data. The arrangements, commercial or otherwise are very critical.
- Within local intelligence, consider connectivity between vehicles, people, and services. There is a lot of work to define what it needs to be. Unlocking 5G offers huge potential. You can connect more vehicles, and to unlock transport networks.
- Strengthen cyber-security: safety has to be paramount. The cyber-security element is coming more to the fore. There is a privacy point of view in terms of sharing transactions and messages. There is a need to ensure cyber-resilience in the face of cyber-threats.
Cracknell concludes that location intelligence is fundamental enabler: “From a technology perspective, it’s how you demonstrate how the vehicle can be safe in different scenarios.” To achieve this connected and automated mobility has to be a collaborative effort.
So, for example, he feels there is a strong impetus for transport authorities and software developers to share information by scanning the environment and by establishing the rules of the road within which local intelligence can operate. These need to be digitized and made accessible to permit the vehicles to understand everything from road markings to speed limits. Subsequently, there is an increasing need for greater collaboration and conversation to establish the paths through local intelligence to fully automated mobility.
