Data Driven Connected Cars as a data-driven ecosystem
The possibilities offered by a Connected Car are very extensive. Consequently, it is by no means just a vehicle networked with the Web, but its data-driven ecosystem. Edge computing, cloud, analytics, and AI play essential roles in this ecosystem.
If you think of a networked car, you also think of the world of smartphones and tablets. Full operability, touch displays, tomorrow's networked technology. Networking offers a lot of improvements, both in terms of comfort and security. Connected cars, for example, can be locked remotely, which saves a second, skeptical step to make sure that the vehicle was parked and secured safely.
But seriously: the advantages are apparent. For example, the car can report itself to its driver when its systems need to be serviced and, in the event of a theft, can announce its current position. Comfort functions such as a tuned suspension system can also be selected, or the chassis can be adapted to the existing road condition - according to the available sensor data. The possibilities offered by a Connected Car are very extensive and can be used in a wide variety of areas.
The vehicle as a data-driven ecosystem
A modern networked car should not only be imagined as a "vehicle with an Internet connection." Instead, it is its data-driven ecosystem that can capture information from the environment itself, including road infrastructure, data from other vehicles as far as they are networked or even pedestrians. This recording of environmental conditions serves not only to increase current road safety but can also help to ensure that the knowledge gained can be incorporated into future vehicle development, road construction, and urban planning. This is made possible by the use of Big Data and Artificial Intelligence (AI), which channel the data streams generated and process them accordingly.
In this context, the term "Edge" is often used. This term comes from the Internet of Things (IoT) environment and means that data from IoT devices is processed where it is generated. The results of this processing are then transmitted to a higher-level computer, which minimizes the need for data transmission. Market research firm IDG predicts that by 2020, investment in edge computing will account for up to 18 percent of total IoT infrastructure spending.
In the automotive industry, edge computing would, therefore, be the processing of all locally generated data generated while driving the vehicle itself. This will enable real-time data analysis without delay, as processing the data at its source will require less Internet bandwidth.
Although Edge Computing has recently become the talk of the town, this technology is not particularly new. Akamai's Content Delivery Network (CDN) developed methods for this type of decentralized data processing over 25 years ago. Edge Computing is usable in various fields. This technology has already found its way into production environments, building management, and offshore oil platforms. These solutions can be used both stationary and mobile, the latter being the case with Connected Car.
Edge computing pioneering in the automotive sector
But why is Edge Computing so interesting for the Connected Car? While driving, a lot of data is collected by the sensors installed in the vehicle. It does not make sense to transfer all this data to a computer in the cloud - the bandwidth is often insufficient. Because intelligent vehicles rely on the mobile network for their data transmission. Some information is simply not relevant for further central processing.
For example, in the case of a parking process in which obstacles are detected while reversing, it is entirely sufficient to process this locally - i.e., in the vehicle - and initiate a braking operation. The situation is different when data is exchanged that enables higher-level processes such as traffic flow control. Here, the individual data collected by the sensors may be less attractive. However, it is helpful for a control center to record, for example, position, speed, distance to the vehicle in front, and ambient temperature to switch the road signs accordingly.
Other vehicles can also conclude their journeys from the collected data of a road user. For example, predictive analyses can help an autonomous vehicle determine when to stop or what to avoid, depending on what might cross its path. Edge Computing can also be used here as a module within an AI system. Using the data collected decentrally, AI models can be further developed independently based on machine learning.
Last but not least, the data collected can also be incorporated into vehicle development. In this way, each driver of another smart car produced participates in the experiences of other users. However, the use of edge computing in connected cars also has another advantage. In addition to saving bandwidth, corresponding solutions can also act autonomously. If the network connection fails, the most important functions can also be performed entirely offline.
Setting the course
Of course, the basis for a meaningful implementation of Edge Computing is a solution that can handle the mass of data correctly and analyze it meaningfully. Without big data tools, the information that is collected, processed, and anonymized during the journey and sent to manufacturers, municipalities, or others cannot be evaluated. IT decision-makers should pay particular attention to performance because the data generated is real-time data streaming in large quantities. However, it is always worth taking a close look at this matter. After all, the analysis of mass data offers companies and organizations added value that could give them a decisive competitive advantage.
This article was first published in German by next-mobility.news.