Dipl. Ing. Joachim Nell, General Manager, CMORE Automotive China
If a car wants to drive alone on the street, it must have a driving license. Though it seems a difficult and unimaginable matter, it is what CMORE has been doing: We design solutions that power computers for automated self-driving cars for manufacturers all around the world.
To become a qualified driver, a person needs a long-term life experience. For example, one needs to understand the traffic signals, familiar with the speed of a car, and have a basic judgment of the traffic condition on the road ahead. But how to teach a computer all about this? Let me start by reflecting on my working experiences.
During the past 3 decades, I almost witnessed every step of the development towards automated driving (AD) technology. It was before 2003 that I was first connected to a driver assistant system. And in 2003, I built a first prototype-car for automated driving, which was not treated seriously at that time when only limited driver assistant-systems were allowed by the industry to help the driver and not to replace him. It was with the launch of Volvo XC 60 in 2008, which was the first car with the so-called “city safety” function that changed the market’s points of view. With the “city safety” function, which is an automatic emergency brake (AEB) system, Volvo XC 60 could automatically brake in the event short before an accident even the driver has not taken any action himself. It was a breakthrough that for the first time a computer can overrule a driver to make a decision. Since then, the ADAS (Advanced Driving Assistant System) has become a major topic, and the whole industry was suddenly starting to rethink about what kind of assistant systems could be created. Above all, the market expectation grew very dramatically. Then, we had a detailed definition for automated driving and the Society of Automotive Engineering (SAE) defined it from Level 1 to Level 5. In L1 and L2, a driver still takes control of a car and only above L3, the driver is allowed not to touch the steering wheel.
Despite that, we still need the driver to take over control again when critical moments happen. We need to use driver monitoring system (DMS) to observe closely the driver’s performance and decide when to interfere and bring him back to responsibility. So, it comes to the birth of CMORE. In order to help self-driving cars with data analysis and monitoring functions in every possible eventuality on the road, a large amount of data collection as well as analysis is indispensable. We provide a chain of data products ranging from C.FRAME, C.LABEL, C.DATA using servers and Cloud, helping our customers to build a comprehensive release recommendation and better meet all requirements from the development of a future mobility system. Our entire complex work can be briefly summarized as the following process: we use the camera or radar or lidar in the car to collect data frames, then it comes to an annotation step, where the C.DATA could provide integration of manual and automated annotation solutions as well as error-free and complete data analysis. By working with CMORE products, our OEMS could get hundreds or thousands of traffic junction situation including the traffic lights, pedestrians, walkways situations to teach the machines by “deep machine learning”, to better fulfill their target of safe automated driving.
In my eyes, the work we have been doing is fundamental and important. A control of a self-driving car can only be realized, when the machine was trained with high-quality Ground-Truth data, telling what is really ongoing in the scene and what has to be detected.
As a young company founded in Germany in 2011, CMORE has experienced a tremendous and strong growth these years. We have been expanding globally, already established branches in HK, Detroit in US and now we are setting up a branch in Shanghai too. In the next couple of years, a team of over 500 people will be built, offering comprehensive engineering and data solutions for automated driving assistance and software development to support ADAS/AD validation in China and globally.
Regarding the industry’s future, I believe that the demand of the mobility would be much more emphasized than ever before thanks to the high car density. Instead of spending more time searching for parking spaces than traveling from A to B, young people will prefer mobility services rather than owning a car themselves, which basically means the decline of the sales numbers of private owned cars and the pop-up of automated cars with functions like AD and ADAS. Therefore, there would be a consolidation of OEMS worldwide and an increasingly call for intelligent engineers since the work would be more complicated and challenging.
Today it’s important to drive a car with a good brand and image to image yourself, but tomorrow, it would be much more valued of the convenience and safety in your trip. If you want to be more efficient and more safe on the roads, we should never under-estimate how important it is to get the right-error free automated driving. By working with CMORE, we can be sure to help you “see more” data no matter what the future holds.