ADASConnected Vehicle

Connected Cars: Innovation Driven by Data

Basic Challenges, High-Level-View

The first steps towards connected cars have been already taken since 2013 with the Connected Car Traffic & Diagnostics Cloud. It uses anonymized traffic movement data and messages and hooks up to traffic participants, road side units, logistic enterprises, OEMS, and others. It is also used for telematic messages for the maintenance of cars and exchanges data with repair shops.

Simon Papel
Product Manager Data-Management, CMORE Automotive GmbH
Simon Papel holds a Master of Science in Business Informatics and has worked for over 12 years for the Deutsche Telekom AG. During that time, he worked on the enterprise critical database for mobile networks storing all information of customers, contracts, products and intelligent network.

Aside to the connection to clouds, connected cars use a heterogeneous communication network between themselves and the surrounding infrastructure with either WLAN-11p (IEEE 1609 802.11p), cellular network communication (LTE-V, LTE/5G) or digital broadcast (DAB+/DMB). The communication network itself must overcome the challenges of connection management, encryption & authentication to offer online services for traffic, parking & toll payment, vehicle sharing, personalized functions on demand, and secure over-the-air-updates.

The communication network has also to share information that is out of sight of the vehicle, like warnings for traffic jams ahead, hazardous locations, road works, lane changes or blocked intersections. Very detailed maps according to the car’s position could be sent on top, when road lines and guard rails are missing in addition to the information that is surrounding the vehicle or the information that is out of sight.

Next to the communication challenges there are legal and strategic challenges:

  • Legal permissions and responsibilities
  • Safety aspect
  • Heterogeneous technical systems of different ages
  • Cooperation of road side unit manufacturers
  • Calculation of signal prediction (latency, quality)

What could be the use cases driven by the innovations of connected cars? Obviously the highly automated and connected driving functions in the context of urban traffic flows and smart traffic control. That also includes co-operative, highly automated and autonomous driving in public transport and commercial vehicles in mixed traffic situations.

For these autonomous vehicles to become a reality, they need data. Big data in fact. The vehicles are furnished with sensors measuring everything from position, speed, direction and braking; to traffic signals, pedestrian proximity and hazards.

The groundwork has already been laid for autonomous vehicles with ADAS functions like collision warning, lane keeping, brake assist, and speed control. In general, all information in these cases is processed locally and kept privately within the vehicle. In contrast, with connected cars, all information can be shared between cars and between cars and any kind of a centralized intelligence instance. By exploiting this shared data, the connected vehicleswill be able to make even better decisions and carry out even more appropriate responses to traffic situations than in the purely isolated case.

Cooperative Driving Functions in Urban Environments

The examples mentioned earlier indicate the potential for innovations that could come with data management for connected cars, but the potential for cooperative driving functions in urban environments are even higher, especially when thinking of a multi-cloud-environment for vehicles, road side units, mobile networks, and a centralized traffic managing system.

·        Vehicular Cloud

The vehicular cloud would be based on temporary vehicular ad hoc networks and clouds to enable the swarm intelligence. Operation of the vehicular cloud be achieved with IEEE 802.11p, 4G D2D and 5G D2D. The major advantage would be that no infrastructure is needed, which makes it especially suitable for emergency cases because of the robustness. From a financial point of view such a solution must be low cost and OEM independent to gain full potential.

·        Road Side Unit Cloud

The road side unit cloud could be based in first step on vehicular ad hoc networks and later migrates to a 5G network. But are the road operators and public authorities ready to operate such a cloud or will there be new market players? The advantages of the road side unit cloud would be, that it taps into the local intelligence for safety critical hotspots and has low latency due to high-grade proximity. Yet who will invest that tremendously when there is a lack of existing business models or operator schemes?

·        Mobile Edge Cloud

The mobile edge computing is the pivotal building block for 4G+/5G networks which is operated by the telecommunication industry, who are firm with 4G (notably LTE-V) and 5G (notably 5G V2X) and WiFi and backbone access technology. The mobile edge cloud can be built on top of existing cellular infrastructures that is then usable for a variety of (also non-traffic) applications and their data. The mobile operator will probably use specific solutions and might even prevent cloud roaming. Having in mind that Europe alone has about 200 carriers for mobile networks shrinking in numbers every day, what will be the chances for new market players? And what about partially suboptimal location of existing cellular nodes for roadside applications? These are all problems that might occur with the mobile edge cloud.

·        Centralized Traffic Management Cloud

All data converges in worldwide or nationwide centralized traffic management clouds created by tech companies, car manufacturers, and service providers having superior storage, processing and communication capacities with the greatest possible economies of scale. Even with latest technologies there are still problems like high latency, security and privacy protection to solve. The centralized traffic management cloud would be a highly attractive target for IT attacks because of the valuable data.

First Steps toward connected cars and automated driving

Connected cars in the context of automated driving need a validation process during development.The development of level 3 to 5 automation systems will become a major technical challenge, especially considering the expected data rates of a level 5 automation system: 17,500 gigabytes per hour. Even a level 3 automation system produces between 1,400 and 2,700 gigabytes per hour.

And this is just the raw data that often needs selection, conversion, annotation for the ground truth, and transmission to software, for example for re-simulation and KPI visualization. That is why data management is so important.For the data management we need to be sure about the entities considered from an ADAS data point-of-view:

  • scenes in every-day traffic,
  • data acquisition,
  • metadata describing the massive data chunks and the data logistics,
  • annotating and labeling data for the ground truth,
  • putting modern sensors in re-processing or re-simulation – including virtual – environments,
  • and create KPI reports out of that.

All these entities are in focus of CMORE Automotive GmbH (CMORE/EC.MOBILITY) and its business unit C.IDS (Integrated Data Solutions). CMORE/EC.MOBILITY is specialized on validating ADAS/AD sensors and meets the current and future data management requirements with the C.IDS development of C.DATA.

Data Management During Development

C.DATA is an automated big data system which highly efficiently and effectivelymanages ADAS/AD mass data along the validation and development process for complex sensors and sensor configurations. The primary goal of C.DATA is to keep manual intervention to a minimum while offering the flexibility and freedom required in a constantly changing data environment. One of the special features of C.DATA is that it is tailored to the ADAS/AD requirements and its continuous validation for level 3 or higher ADAS/AD functions. Everything from scene management, vehicle construction, data logistics, annotation data, re-simulation to KPI management and KPI reporting is combined in one platform.

In addition to C.DATA, CMORE/EC.MOBILITYdeveloped the PODBOX (Persistent Onboard Diagnostic BOX). PODBOX as a multifunctional platform offers custom measurement technology solutions for use in the vehicle or in the laboratory and additionally for prototyping algorithms and functions. It is a multifunctional measurement and diagnostic platform for intelligent data analysis. It enables remote access to ECU diagnostics including data logging (CAN, video, GPS, online tagging, etc.), automated tests and reports, system monitoring, and documentation of test drives. It integrates over-the-air functionality (WiFi, 3G/GSM) and provides real-time data in multiple views.


Think of all the possibilities and rewards when harnessing the data of connected cars and the disruptive potential for innovations regarding future mobility. Travelling overall will become a much different experience: safer and more flexible with the seamless integration of personalized data. Safer for all participants of every-day-traffic – even when not in sight – including new mobility concepts for commercial and public transport.

The data management to enable all that should not be seen as hinderance or burden, but as a chance to uplift modern – and much more mobile – society.

About the Author

Simon Papel

Product Manager Data-Management, CMORE Automotive GmbH

Simon Papel holds a Master of Science in Business Informatics and has worked for over 12 years for the Deutsche Telekom AG. During that time, he worked on the enterprise critical database for mobile networks storing all information of customers, contracts, products and intelligent network.

Back to top button