Specialist Patent Attorney and Automotive Safety Specialist Simon Parry considers the not-too-distant future of automotive tech in the UK.
On 18th August 2020, the UK government’s Department for Transport (DfT) issued a call for evidence on the use of Automated Lane Keeping Systems (ALKS), which marks a tentative but positive step towards automated driving technology being permitted for use on the UK’s roads, and which could see the first such systems hitting the roads as early as Spring 2021. But what is this about, why is it significant, and is the technology sufficiently developed for this to be sensible?
Many readers may already have encountered so-called Advanced Driver Assistance Systems (ADAS) and likely even drive cars equipped with such systems themselves as ASAS is already permitted on UK roads. ALKS would be a significant step forwards from ADAS.
ADAS systems include technologies such as advanced emergency braking and adaptive cruise control, but also (and not to be confused with ALKS), lane keeping assistance. Importantly, ADAS systems are classified as Level 1 or Level 2 systems under the UNECE’s 5-level classification system of driving automation, where Level 5 denotes fully automated driving. Level 1 systems function to provide either steering or braking/acceleration assistance to the driver but not both together, whereas Level 2 systems provide both steering and braking/acceleration assistance together (for example via the simultaneous operation of adaptive cruise control and lane keeping assistance). Importantly, Level 1 and Level 2 systems provide driver support features only, and are not considered to provide automated driving because both require the driver to remain in overall control of the vehicle and to constantly supervise the systems to ensure their safe operation.
This is where ALKS systems go a stage further, because the ALKS technology on which the DfT consultation is focussed operates at a significantly higher level than ADAS. ALKS systems are classified as Level 3 systems providing actual automation, and do not require the same degree of oversight by the driver. This is a very significant jump, and will focus considerable interest on the DfT’s consultation, because moving from Level 2 to Level 3 requires acknowledgement that the vehicle itself will be taking over actual control from the driver, albeit for limited periods and only in particular environments, with the driver required to take back control if requested by the system.
ALKS systems will be required to permit their operation only if they deem the driver to be available to take back control if called upon, and must actively monitor the driver’s availability throughout their operation. They must also be configured to follow a series of escalating alerts to an inattentive driver. The systems will also be required to bring the vehicle to a safe and gradual stop in its lane if the driver fails to respond, and to perform an appropriate emergency manoeuvre if it detects an imminent collision risk. This will be a big change from how ADAS systems operate.
The move from a driver support system (ADAS) to an automated control system (ALKS) raises some potentially thorny legal issues regarding who is responsible in the event that an accident occurs whilst the system is in operation. Should the driver be held accountable, or the system provider, and under what circumstances might that liability shift from one to the other? The DfT consultation seems like to generate lively debate and diverging views on this, and will be observed with interest by the industry because the arguments advanced either way could have far-reaching consequences as systems begin to develop further towards Level 4 automation, where the driver will not be required to be on hand to take back control.
Because ALKS systems take over the actual control of a vehicle, they are required to maintain very accurate situational awareness, including not only the position and trajectory of the host vehicle, but also the position, current and predicted trajectory of all other vehicles in the vicinity of the host vehicle, as well as accurate information regarding the road ahead. We see ALKS systems using many sensors of various types, RADAR, LIDAR, cameras, and also vehicle-vehicle communications, and vehicle-toinfrastructure. Even more sensors are used to adequately monitor the driver to ensure alertness, by tracking eye and head movements. All of these sensors generate a very large amount of data, much of which will be required to be stored for analysis by the relevant authorities in the event of an accident. Cybersecurity is therefore expected to be another key area of debate in the consultation.
For a government keen to put the UK at the forefront of advances towards fully automated driving, this consultation on the potential early adoption of ALKS technology seems timely given the current state of the art in ALKS systems. We have seen from extensive patent filings that the technology has been developing very significantly over a number of years now. ALKS is being proposed only for use in the relatively controlled environment of motorways, which have relatively regular flows of traffic moving in a generally uniform direction and following clearly defined lanes, and where pedestrians are prohibited. This fits well with the current capabilities of these types of systems and the focus of innovation which we have seen in this area of automation. We have seen particularly interesting developments in the computation of lane trajectories in the absence of clear road markings, the tracking of other vehicles, and the use of clustering techniques to supplement lane trajectory predictions using traffic flows.
The DfT’s consultation is due to close on 26 October 2020, and details of the responses are awaited with interest. Might we actually see this very significant handover of control take place next year? We will have to wait and see.
This article was originally publsihed in UK Tech News. View the original article here.
.png?width=100&height=100&name=Simon%20Parry%20circle%20(4).png)
