As part of the Continental Commercial Vehicle Days online conference, one of the Web Sessions was focused on environmental detection practices for today’s market, as well as the general challenges faced by the commercial vehicle sector.
Continental’s Online Conference Focuses on Environmental Detection
As noted by presenter Christopher Schrecke, Head of External Communications for Commercial Vehicles and Services, Continental, safety remains a significant challenge for the industry with many drivers turning in the wrong direction at the wrong moment, causing a major, sometimes fatal safety incident. With an example cited of over 4,700 accidents recorded in the US in 2017 – an average of around 12 truck accidents a day, this session offered a chance to address these concerns. Handing over to Fässler, the talk addressed the pivotal role safety would play in the future evolution of commercial vehicles, in line with the implementation of the EU General Safety regulation, planned for 2024.
The talk, led by Georg Fässler, the Head of Business Segment Products and Systems, Continental, started by his briefly explaining the rapid technological changes in the industry, aided by AI and computation. Fässler believed there were four main challenges being faced today. Lack of truck drivers who are unwilling to drive such long distances is one major element; low margins of fleets is another, meaning that the costs need to be minimised as much as possible from A to B; thirdly, the high number of fatalities; and fourthly, a more urgent need for sustainable transportation.
Fässler mentioned that perception-based functions – which would be added both inside and around the truck, would allow for much more proactive solutions for the aforementioned challenges, specifically for fleets, by improving fuel-consumption and maintaining a better ecological footprint and offering an extra level of protection within society. Eventually the perception model should allow fleets to have more driverless vehicles, which will help reduce costs for fleets in the long run.
The perception components around the truck will help give a 360° perception, helping to detect and classify objects – be it pedestrians, cars or bicycles. There’s also a possibility to detect perception of the objects or persons, via their motion (or eyes), so if they’re waiting at the pavement or intending to cross the road, for example. Combining those sensors in an overall system is a major development as it allows for all directions, which means perception both inside and out the vehicle, as well as beyond the possible sight lines of the driver. This adds a degree of responsibility for the driver, as fleet managers can be aware if their driver is paying attention to the potential dangers that come with driving, such as icy road conditions, or vehicles or objects within the sight line.
The automatic emergency brake assist system, as part of the forward-looking perception technology available, has two core aspects; a camera and a radar. The radar has two beams, one with a long range and one with a shorter range. The emergency brake assist works based on what can be seen in the camera view and is designed to fully stop within a distance of 200 meters. In the future, the plan is to incorporate a speed assist function, based on looking at the various speed limits, and this will eventually go into the perception management package. Congestion will also be crucial, as the information system can help inform actions within urban settings and will be mandatory for 2024. Equally, crucial, side looking perception will help in detecting blind spots, especially when looking out for cyclists, which can be missed by human error.
Inside the truck, there is a detection function for drowsiness, which helps measure the driver’s alertness. The LED lights will help give enhanced views of the surroundings. Crucially, to enhance safety, there will also be medical checks included, including hints on how to improve on driving techniques.
Another new development in the future will be eHorizon technology, known as perception beyond the sight of the line. Dynamic changes will be made with eHorizon, which helps detect different problems drivers face, such as icy conditions and curves and hills ahead. Also offered with eHorizon is the opportunity to see traffic jams ahead of you. One of the key elements of eHorizon is its ability to store MAT (Moving Annual Total) data. In the future, more dynamic data will be available, helping you to decide when to slow down speed if there are difficult roads ahead. This is especially useful for drivers with loads of 40 or 50 tonnes and will help in saving fuel.
Continental is investing in this area by testing for autonomous driving on the motorways of Frankfurt. The tests include how to overtake cars and enter and leave motorways. In the future, this will lead to driverless manoeuvring in confined areas, which will be easier to navigate than urban centres. The final plan is to bring more autonomous driving to the motorways after extensive testing.
A variety of questions were posed after the webinar, such as how well smaller objects would be detected with the sensors. Fässler conceded that more sensors and robust localisation would help encounter those objects, but it required a huge amount of energy to introduce. With large trucks, compensation would be needed for nicks and rolls of the cabin, which would be provided by self-calibrating cameras and radars, whilst the largest truck models would also require wireless transmissions protocols from the back of the trailer to the front. This is a crucial topic as the latency time to get the information of the sensors from the back to the front must be within 200 milliseconds.
With EU legalisation, the ultimate Continental goal is to have driverless engines, which is classed as the highest level of automation. Core elements would be needed to achieve this, such as leading-edge radars, algorithms, solutions in line with GSR (General Safety Requirements) and telematic solutions, via eHorizon. All of these safety measures would aid the fleet in making good decisions based on where to drive and what to do in the truck. If the automation technology failed, they would have an installed control unit backup.
One of the questions that will eventually be answered in the future would be the current connectivity question, would Wifi car to car, or 5G coverage prevail as a unique system of data coverage? Fässler asserted here that he believed the technology, in line with eHorizon, would converge to 5G, emphasising that a strong data connection was an essential component to the technology functioning seamlessly.
Fässler concluded by acknowledging that he imagined that the majority of the applications referenced in the webinar would more than likely be utilised for usage on trucks, especially as the majority had been legally mandated. Having started the discussions with OE manufacturers, Fässler believes the functions will come step by step and will be enhanced as the progress gains more traction in the future. Not only that but the technologies would directly help reduce fatalities on the road, with Fässler citing the radio detector system as being like a second driver, looking over your back as a means of additional support.
With viewers asking questions via Continental’s Slido there was some interesting discussions away from the webinar. When asked if Europe was a ‘pioneer’ in safety leglisation, Fässler mentioned it’s more to do with safety legislation, whereas in the US, it’s more driven by insurance, to avoid damage costs at the end, however China is now moving towards introducing legislation, which shows that the “Whole world is becoming more and more interested.”
The most interesting question asked on Continental’s Slido function, was how Continental would deal with the legal ramifications, or the ‘blame’ if an accident occurred in connection with the environmental detection system. Fässler argued this question philosophically, explaining that this is a relatively new development and is still ‘in production,’ which sums up the continuing evolution we can come to expect from environmental detection in the future.