Much has been written about over-the-air (OTA) updates reducing the need for recalls of essential automotive software, but there are many other use cases for OTA updates. There are two aspects of OTA updates that need expounding: What is being updated and where it is being updated.
Based on our experience at HARMAN and with our customers, we can see that while software updates are first and foremost being deployed OTA to the head unit and the TCU (modem), the ECUs that are gaining the most attention in car model 2018 and higher are the ADAS ECUs. Be it the camera, radar, laser or ADAS aggregator, ADAS is the functionality that is the most complex and has the highest time-to-market pressure, which can lead to software problems that will benefit from OTA updates.
ADAS is not only software driven, but also very dependent on the quality of the HD maps that run it. These maps will get more detailed as ADAS enabled cars crowdsource their information and the roads we drive on change at an incredible rate, considering construction and the effects of extreme weather --all pointing to the need for map updates. Updating maps as close to real-time as possible will be critical.
As a side-note, updating maps OTA is an entirely different beast than updating software. The format of maps is fundamentally different from that of embedded software and HD maps involve many gigabytes of data.
OTA software management brings many safety and infotainment benefits to the driver and also enable manufacturing efficiencies for the OEM. One set of hardware can be deployed for multiple car models shipping to different regions with the difference being in the configuration. For example, OEMs can read a car’s configuration file and update it over-the-air.
And now we get to where the updates happen. Sure, the most disruptive use of OTA is to perform updates to the car while it is on the car owner’s property, removing the need for a recall, an approach we see with Elon Musk and Tesla. However, OTA brings efficiencies and benefits to at least two other stages of the car’s lifecycle. Take for example the production line. The many megabytes of code that are loaded onto the dozens of ECUs in the modern car will require a big flashing window on the production line to complete if it is done in the traditionally serial production line manner. With OTA, once the battery and the TCU are connected to the internal CAN BUS, any ECU that is connected can be updated and flashed on the fly – even if, for example, the current station is the one for the installing the upholstery. Not only would this OTA approach remove the need to add costly time to the production line and remove the risk of software-related bottlenecks, but it could also lead to production line efficiencies by reducing the time currently allocated for software flashing.
Another element to consider with regard to the growing prevalence of software in the car is the traditional skillset of the local garages/dealerships. Training the hundreds of thousands of mechanics around the world to understand, diagnose and fix software-related issues is a task wrought with challenges where the cost of mistakes and failure is high. Centralizing the knowledge base to regionalized centers of excellence will enable a highly trained staff to remotely manage cars OTA while they are at the garage. In this way, the garage can focus on performing the maintenance tasks and mechanical fixes required, while in parallel the car is diagnosed and updated as it is connected to the garage Wi-Fi network.
The use cases for OTA are vast and varied, enabling every OEM and every region to roll out update services for different use cases – if not all of them – from day one.