How does edge computing benefit connected cars?

Edge computing has gone from being a buzzword to become an increasingly standardized technology; one with great potential, not least in the automotive industry. But what is it, and how does it contribute to the development of connected cars and digital car services?

What is edge computing?

Edge computing is a distributed form of computing. This means that many computing operations are performed on a device, rather than in the cloud, with the edge in question referring to the outer edge of a cloud network. The purpose of this is to bring data collection, data analysis, artificial intelligence, and other computing resources closer to where the data is collected and presented – such as in a car.

How is edge computing useful to car makers?

Outlined below are some of the many ways in which edge computing can be useful to car makers:

Decoupling software from hardware (and the shortening of development cycles)
Edge computing allows software to be decoupled from hardware, meaning that the development of services can be detached from the development of physical equipment. As software development cycles are shorter than hardware development cycles, it gives car makers and service providers a whole new kind of flexibility. Services can be added and updated even after the car has been sold. Traditional over-the-air software updates in automotive have targeted the update of the complete firmware in one or more ECUs (electronic control units). With edge computing, discrete function updates can be provided to individual cars and their ECUs. It is a principle that can be compared to updating individual apps on your smartphone versus updating its entire operating system.

Increased local processing power
Edge computing in cars will be required to handle the exponential growth of data, as the number of sensors will grow when cars get more and more intelligent. As edge computing is a decentralized extension of the computing cloud, limitations in terms of car computing power and network capacity can be circumvented. When sufficient computing power is available, this opens up the potential for machine learning and artificial intelligence, as well as a smarter way of handling situations when the car is offline. For example, if there is no internet connectivity during an emergency call, but voice call is possible, the edge code can audibly announce the vehicle location to the call center agent over the voice call, instead of sending it as a data message.

Secure management of safety critical applications
Having edge devices built into a car ensures that it will receive the information it needs in order to drive safely, even when it is offline. No car can be expected to be connected to the Internet at all times. Safety critical applications can only be fully relied upon when their data is processed accurately and quickly within the car. Edge computing can play a vital part in ensuring this.

Bridging the divide between on-board and off-board
Edge computing bridges the gap between embedded in-vehicle (on-board) software and outside-vehicle (off-board) software. This makes it easier for cross-functional teams to design services end-to-end, and maintain focus on a holistic end-user experience. This, in turn, is very valuable to car makers as they develop new services. 

Using multiple software suppliers requires a customized cloud solution

All major cloud vendors – Amazon Web Services, Microsoft’s Azure, Google Cloud Platform – have their own preferences when it comes to edge computing, and the distribution of software to “the edge”. While car makers always select a single cloud vendor for their edge computing needs, it is not unrealistic to expect that the vehicles of the future will contain edge computing technology from several vendors. However, the established cloud vendors of today do not offer standardized solutions for certifying the software used in edge computing. To use another smartphone analogy, there is simply no “app store” from which approved software can be selected and implemented. Using multiple software suppliers therefore requires a customized cloud solution.

How to realize successful edge computing services in connected cars

What are the most important factors for car makers to consider in order to make edge computing services for connected cars convenient and reliable?

• Teams having end-to-end responsibility for their services when developing and operating them. The more a certain team can be responsible for a certain service, all the way from the car to the cloud, the better. Involving a great number of teams, each one responsible for a certain layer of the overall service, is less efficient and reliable.

• Having a service platform or process that allows for services from multiple software suppliers. Again, similar to an app store, where the services have been reviewed and approved. This kind of efficient general solution has yet to be realized, but meanwhile each car maker can manage their own solution. 

• Having a clear, standardized vehicle API for the communication with the car from the edge software. The better the API, the better the services. Updates will be easier to implement, and response times can be shortened.

Edge computing is the way of the future for connected car services

Edge computing, not least in the world of automotive services, is something that we at WirelessCar believe is only in its nascent stage. It is a technology that should be of great interest to all car makers, and one that we invest a lot of time and resources in.

One should expect edge-to-cloud capabilities in vehicles to become more and more common, as this is a natural step in the process toward greater car connectivity. At the time of writing, we are working on edge computing projects for several car makers. Our services are built to integrate with any edge computing platform.

Do you want to know more about how we work with cloud and edge computing, and what applications we can build for your company? Feel free to contact me at Tomas Carlfalk. You can find out more about how we work, and our WirelessCar cloud strategy, here.