Working group on vehicle localisation
András Csepinszky, head of the working group on vehicle localisation
He wanted to play, but no games were available
András Csepinszky is the director of advanced automotive technology at NNG Llc. started as a biochemist. His father was a teacher at a college of agricultural studies in Mascara when he graduated from the Pannonhalma boarding school. His father invited him to Algeria, and he went with the thought of “at least I will learn French.” That is where he started his studies of Biology, which he carried on in Debrecen and finished at the prestigious Maitrise de Biochimie, Louis Pasteur University, Strasbourg, France. When he returned to Hungary in 1995, he could not find a position matching his expertise. For a while, he was working as a translator, and in the meantime, he studied information technology.
He first started writing computer programs in the early ’80s as a child at a middle school in Nagyváthi. His godfather was teaching Mathematics and Physics. Hence he could use the computer at the school. He only wanted to play, but no games were available on the device, so he wrote one. In 2006 he was already fully committed to vehicle telematics at the Connexis Kft. in Debrecen. From 2009 he was a project manager at ERTICO-ITS Europe in Brussels for more than five years. In 2014 he created the company-wide global Project Management Office of NNG Llc.
András Csepinszky is 50 years old, married with six children.
For us, “Karesz” is the number one priority right now.
Vehicle localisation is not a new science or technology. The autonomous vehicle technologies did not create it, but they sure made it very important. So much so that the Mobility Platform of KTI set up a working group on vehicle localisation. But why exactly? That is what we discussed with the head of the working group, Mr. András Csepinszky.
-What does vehicle localisation really stand for in the name of the working group?
-This is not just some new science. It was not brought to life by developed vehicle technologies or by the development of autonomous technology; however, many people might imagine that. Ever since navigation exists, we must be able to locate the exact position of a vehicle. It has to be able to be shown on a map where a given vehicle is at exactly since if we have no information on its position, we cannot navigate it anywhere.
-What does “the exact position” actually mean in this context?
-It all depends on the purpose of why we would like to locate the position of a certain vehicle. Different level of speed and accuracy is required when the vehicle is driven by a human driver, who is simply trying to navigate it by using a GPS navigation device. In this case, it won’t be a problem if the GPS service provides users with approximately 10-20 meter accuracy since a human driver can correct such inaccuracy. The human driver can see it if the street where he/she would like to turn is a little further down from where the GPS device indicated, and he/she will act accordingly. However, it is a different situation when the vehicle is driven together by both humans and machines or by only a computer. For the latter, the position of the vehicle must be determined with a centimetre-level precise point positioning.
-It seems that the accuracy of the navigation devices significantly increased. What has changed? The technology, the software, the satellites?
-The technology has already been available for decades. With the military application of GPS signals, there have long been certain systems that could determine even the position of a moving object, with a centimetre-level precise point positioning. However, these military applications could not be used for civil purposes.
-What are the challenges of vehicle localization?
-Nothing and everything. We must be able to provide answers to certain questions, like what information is required for the implementation of driverless technology? What is the requested level of sophistication of the purpose-built HD Map dataset available in the memory of the vehicle? What are the quality and the quantity of data, which the driverless system should receive from the satellites, what information should it collect by using its sensors, with a sufficient degree of speed and accuracy? What data has to be provided for the driverless vehicle about weather and traffic situations to ensure safe operation and how will the driverless technology, hence the vehicle itself, recognise a slippery surface, for instance? I need hardly say, if the sensors detect that the wheels are slipping, it is already too late….
-Meanwhile, I assume that the gear shifting of the driverless vehicle would not rely solely on the reports of the Weather Information Services
-It won’t rely entirely on their reports – it’s more accurate to put it that way. The driverless vehicle will be aware of the temperature is -5 Celsius degree when the roads are expected to be slippery. The sensors of the vehicle see the surface of the road, of which they can draw certain conclusions about its condition and how speed and breaking can affect road performance. Moreover, other vehicles, going in front of it, can signal if they ran onto a slippery surface.
-When will all this become a reality?
-I obviously cannot give you an exact time. In any event, the European Commission already appointed a group of experts to prepare the legal framework, which will have to be implemented in all Member States.
-What are the major tasks of the Working Group on Vehicle Localisation within the Mobility Platform?
-Our task is a type of applied research, to put it simply, it has to find an answer to the question of what shall be done in Hungary for achieving operational driverless technologies. The first and currently also the most important question to which we try to find an answer is whether the Public Road Data Collection System of Budapest Közút Ltd. (KARESZ) able to provide a solid base to the map, according to which the driverless vehicles will be able to navigate. Budapest Közút developed KARESZ for public transport purposes; meanwhile, if with proper funding, it can be further developed for more widespread use, it can be highly advantageous for the entire country.
-Most importantly, we would not rely on foreign map providers… The information used by driverless technology would remain a Hungarian data treasure.
-However, it is common knowledge that in Hungary, the actual positions of certain streets or buildings can be located elsewhere on maps…
-It is unfortunately, true. That is why we came up with the idea that perhaps the driverless technology could also be used for upgrading maps. However, it is even more important to find out if KARESZ is fit for the purpose, and I must emphasize that this is still subject to more detailed assessments. If so, internationally accepted standards could be used for managing all collected data, hence all vehicles, equipped with driverless technology, and placed in circulation in any Member State of the EU, be able to communicate with the system.
Fekete Gy. Attila
Photo: Iró Zoltán