Positioning as a Cloud Solution for ISA

Digital speed assistants (ISA) need to feed in feedback when the applicable speed is exceeded. With a cloud solution, the data basis for this is accurate.

Modern cars are increasingly becoming driving computers that monitor the vehicle's environment, the driver and his driving behavior: systems that, for example, directly acknowledge a lane change without a set turn signal with a vibrating steering wheel, detect tired eyes or continuously monitor the distance to the vehicle in front. The Intelligent Speed Assistant, which will be mandatory for new vehicle models in the EU from 2022, is also included as a protective measure. All these systems are designed to reduce the number of accidents and make driving safer overall. Such systems are already in use today - without being mandatory - and give a taste of what is to come.

Numerous drivers have had their first experience with these systems and observed that it is not always easy to implement speed assistance systems without errors: the instrument cluster in the car shows 30 km/h, even though the driver made a sharp left turn just before entering the 30 km/h zone, where the speed limit does not apply. Or the driver is surprised to see that the speed limits displayed on the same route in the opposite direction are different from those in the current direction of travel, even though the same roads are being used. On the freeway, a speed limit is not recognized because, due to weather conditions and the right lane being completely blocked by trucks, the traffic signs cannot be read properly. Annoyingly, the instrument cluster shows 120 km/h, although only 80 km/h are allowed -?when?a short, warming flash then often ends the?assumed reliability and safety of the system.

You don't have to be a software engineer to imagine what else is possible: What would happen if the car incorrectly recognized the maximum speed on traffic signs? For example, if you're driving at 120 km/h on a city highway and the built-in cameras detect a 30 km/h sign. Is this false information possibly evaluated by the vehicle and initiates a dangerous braking maneuver?that the drivers behind you do not expect?

Cameras provide the basic information for ISA

All systems rely on smart cameras and thus on those little helpers that not only capture an image, but are also equipped with the intelligence to evaluate the information supplied according to a programmed scheme. These detect the maximum permissible speed by optically registering the signs and pass this information on to the driver on the display; if the speed is exceeded, the system signals this with a sound signal or an overlay in the instrument cluster. This generally works quite well, but of course fails when there is no clear view of a sign.

Intelligent speed assistance and digital maps

Camera-based systems alone are therefore not sufficient to reliably determine the maximum permitted speed. In line with the motto "redundancy creates safety," it makes sense to support the cameras or intelligent driving assistance with independent components. One obvious solution is to use digital maps in the vehicle. A navigation system quite naturally uses a digital map. Such a map typically knows the speed limits of roads, and the navigation system knows exactly where the vehicle is on the map. Thus, the current speed limit is also available in the navigation and can be used together or in addition?to?the information from the camera systems to achieve more accurate results.

A prerequisite for the reliable functioning of the digital map is, of course, that the available data is up-to-date. Given the update cycles, even if they could be OTA (over the air), there is of course always the possibility of unknown changes. This is even more true when navigation systems have to be updated via CD, DVD or USB.

Ensure ISA function even without navigation system

However, not every vehicle has a navigation system. What will happen to the low-priced cars that do not have a navigation system by 2024 at the latest, when no vehicle will be allowed to roll off the production line without an ISA? In addition to vehicle navigation, modern cars use solutions that determine an electronic horizon independently of navigation, which is also based on a digital map. However, unlike navigation, the function of such an electronic horizon is not directly visible to the driver.

Nevertheless, with such a solution, the vehicle position must be determined on a digital map. The characteristics of the road currently being traveled are available at the end, as with a navigation system, which means that this solution can also provide the maximum permitted speed. The determination of the position on which road of the digital map the vehicle is moving takes place in the vehicle. Both navigation systems and electronic horizons place demands on the hardware in the vehicle. The digital map has to be stored, and the position determination puts a load on the CPU, which is typically busy with many different tasks. If the map in the vehicle covers all of Europe, for example, then 15 GBytes of system memory may well be occupied. Additional memory is needed to update these maps, because the "old" map should be usable until a new map is available. And regardless of the technical possibility of receiving OTA updates, it is mandatory that the vehicle has at least a GPS positioning system.

Cloud-based positioning as an additional safeguard for ISA

Neusoft offers a solution where a digital map in the vehicle is not mandatory. This also eliminates the need for a sometimes time-consuming?updating?of the map. The digital map is held centrally in the cloud and updated there in a timely manner. Map matching also takes place in the cloud, while the GPS or dead reckoning positions are recorded in the vehicle. This allows the load on the CPU in the vehicle to be reduced.

There is a large degree of flexibility in the design of the cloud solution: depending on the wishes of the car manufacturer or a Tier-1, Neusoft can provide the complete cloud solution to retrieve the permitted speed directly from the Neusoft cloud; however, the speed assistant solution can in principle also be integrated directly into any other cloud with?little?effort.

This cloud service is able to process a short position track plus optional additional data to map it onto a digital map. In the vehicle, this only requires a GPS receiver (in the best case with integrated dead-reckoning) and an Internet connection. Minimal logic that determines the position track and then sends it to the cloud service and processes the response in the vehicle is also required. A complete positioning solution, including a local database or connection to a database in the cloud, is then no longer required in the vehicle.

In the vehicle, this requires a telecommunications box (T-box) that is responsible for the connection to the?internet, whereby a?client of the cloud could even?already be integrated in the T-box. In addition, a GPS receiver (preferably with integrated dead reckoning) is needed to determine and collect the vehicle positions and package them in a cloud request. In the vehicle, the system then needs?software to send this data to the service as a response to request?and, as a result, subsequently extract the appropriate speed from the responses.

ISA always up to date - with low data volume

The amount of data exchanged for this is comparatively small, which is why there are no special requirements for the bandwidth of the connection. Much less than 1 KByte is required per request,?and as a result the amount of data is even smaller, since in the simplest case only the speed is delivered. In the end, the connection quality determines the speed of the information received. The processing times in the cloud are negligible compared to the latencies during transmission.

A key advantage of this solution is?timelines. The digital map is centrally located in the cloud and can be updated at any time as needed. Changes to the road network can be responded to immediately, so that all vehicles can always be provided with the latest information. Theoretically, it is possible to react immediately to a change in road routing?as a response to road works, and temporary speed limits can also be taken into account very promptly, provided they have been made available and updated accordingly in the digital map. Provided that the digital map is up-to-date, the result is always up-to-date from a purely technical point of view.

The user is free of responsibility with data updating automatically.?Also, because map matching takes place in the cloud, there is no need to load large amounts of data from the cloud, which would be necessary with in-vehicle map matching.

How the cloud responds

What happens to a request in the cloud? Unlike in-vehicle positioning, the request comes virtually out of nowhere. There is no history whatsoever to help resolve the ambiguity of positions. Nevertheless, such a situation is not completely unknown for the positioning of a navigation. It is roughly comparable to the situation when a vehicle arrives on the road network from off-road or even from a parking garage. There is no evaluable history, so the position has to be mapped to the road network without knowing where it comes from. This so-called "stateless map matching" is already an integral part of elaborate positioning solutions, but with the difference that this function was previously at home in the vehicle and not in the cloud.

Map matching for ISA

Here it is possible to benefit from decades of experience in the vehicle in the cloud as well. Problem scenarios are well known from many years and have been solved. However, experience also teaches that stateless map matching without a history can at most be as good as stateful map matching. Therefore, in stateless map matching in the cloud, a short position trace is used for mapping to the digital map. The trace can be enriched with additional data as needed to resolve ambiguities. Such ambiguities occur especially when parallel roads are very close to each other. For example, the speed can be additionally used to plausibilize a position. However, even such a solution cannot be perfect.

Optimal solution for ISA: camera, navigation and cloud map matching

For an optimal solution, the smart camera, integrated navigation and cloud map matching play together. The camera always delivers up-to-date images and plays out its advantages when visibility is at its best, while database-based solutions offer advantages in unfavorable weather conditions and when traffic signs are shaded. The cloud solution scores particularly well with its?timelines?in cases where the camera fails. As a minimal setup, the cloud solution can be the ideal complement to a camera-based system. If a navigation system is already on board, the cloud solution as a small supplement still offers the advantage of being able to respond better to changes in the road network in case of doubt. For example, recent road works and changed lane layouts could quickly overwhelm the on-board navigation system. The navigation system still notices that the vehicle is not moving in line with the map, but cannot do anything about it. In such a case, it could send a request to the cloud, which already knows the changed lanes, takes them into account and therefore maps the vehicle correctly onto the road and determines the current maximum speed.

Of course, it is possible to ask the cloud exclusively in case of need. It requires almost no hardware resources in the vehicle, which is especially beneficial for vehicles without navigation. This creates safety in the service of the driver through redundancy. A positioning solution as a supplement from the cloud makes intelligent speed assistance significantly better without burdening the scarce system resources in the vehicle.