Enabling Innovation Through a Digital Feedback Loop

If your smartphone could never download new apps or receive software updates, how quickly would it start to feel dumb? Most consumers would not be okay with that kind of stagnation. Yet, we have traditionally accepted a lack of feature freshness from the most expensive "devices” in our lives — our vehicles. Fortunately, that is changing quickly.

A digital feedback loop (DFL) provides the two-way communication needed to enable such ongoing innovation. Every button push, voice command and user experience provides critical insight into what features customers love and which need improvement.

With a DFL, vehicles collect data about the applications’ performance and use and send it to the cloud so data scientists can analyze it to identify trends and uncover unmet customer needs. Developers can then use those insights to create improvements that are pushed to vehicles through over-the-air updates, and the vehicles once again collect data about the application’s performance — thus completing the loop.

Highlighting the power of DFL, Aptiv and Wind River are using Wind River Studio to create innovative new features and release them through over-the-air updates to test vehicles that are already on the road. Despite being on a much smaller scale, this same functionality and automation can be applied to OEM fleets.

As with all of the best innovations, this one starts with an insight.

Turning data into insights

Imagine a busy parent with their arms full of groceries approaching the trunk of their vehicle. Their smartphone is equipped with a phone-as-a-key app, but they cannot reach it. So they fumble over themselves trying to activate the kick sensor that opens the vehicle’s trunk.

On its own, this moment represents a single data point, but when it is transmitted to the cloud and aggregated with similar moments experienced by other users, a trend emerges that provides insight into a common user behavior: When a user stands behind their car for three or more seconds, they are usually trying to open the trunk.

Develop, test and deploy

Armed with this information, a developer can write a new piece of code for the phone-as-a-key app that controls the vehicle’s trunk so that it opens automatically when a user stands behind their vehicle with their phone for at least three seconds.

Saving the new code pushes it into a pipeline in Wind River Studio, where the developer can scan for defects using a wide range of third-party plug-ins, such as Coverity. They can also conduct tests on virtual targets — simulated versions of the real hardware — to see how the software change would behave in real life. Reducing developers’ dependency on physical hardware accelerates the process and reduces costs.

After the code passes the pipeline’s automated tests, the update is cleared for release. From Studio, the developer can push the new functionality over the air to the trunk app. And just like that, the OEM has added a new feature to improve the consumers’ user experience.

Putting the digital feedback loop behind a single pane of glass

Vehicles are the most complex devices in our day-to-day lives, with dozens of features and functions that must work seamlessly to ensure not just a safe drive but also a pleasant user experience. The hundreds of hours that consumers spend driving a vehicle each year can generate invaluable data.

In the example above, we analyzed the position of the driver outside the vehicle, but the same principles can be used to analyze any piece of data from any device — including even far more complex operations, such as ADAS features, in-cabin user experience and propulsion systems.

Wind River Studio makes the entire digital feedback loop accessible through a single pane of glass, providing all of the tools developers need to aggregate data from thousands of vehicles in the field, write new code to address problems, or create entirely new features.