A wild battle is raging in the car between Google, Apple and automobile manufacturers! Google and Apple are attempting to take over the automobile interior with Android Auto and CarPlay. While the OEMs are still playing along right now, they are also searching for more attractive solutions to connect the smartphone and car. Technologies such as mySPIN, Smart Device Link and MirrorLink are intended to enable this. We tested them and also present the newcomer OpenCar.
Three different main philosophies can be identified:
This is where the different cultures and interests of the app developers on the one hand and the automobile industry on the other hand encounter each other. An exciting clash of cultures!
There are a few exciting technologies aside from the top dogs Apple and Google, some of which have been in the market far longer. Here one can clearly observe how different technologies are developed, tested and in part discarded again.
In this article, we are sharing our experiences with MirrorLink, mySPIN, Smart Device Link and OpenCar. We briefly explain the technical concepts and our experiences with development and use in the vehicle.
MirrorLink is the product of the CarConnectivity Consortium of some OEMs, including VW, Honda and Peugeot.
MirrorLink is based on a technology that is also used in remote desktop software. The screen contents of the smartphone are transmitted to the car via a cable, touch screen input is transferred back to the smartphone. MirrorLink apps have to be reviewed and approved by authorised bodies before they can be used in the car. Once an app has crossed this hurdle, it can be used in all MirrorLink-capable vehicles. Unfortunately MirrorLink can only be used with select Android smartphones.
First of all, the app must have a landscape mode and be displayed in this mode as soon as it uses MirrorLink. Only a few modifications are required in the app itself. Integration of a MirrorLink library at a central location in the app is one of these. It then looks after transmitting the screen contents to the vehicle. The app also has to contain a MirrorLink certificate that identifies it as an officially approved app. This certificate has to be created again by the developer for each new version of the app.
Implementation is effortless once you know how this is done. Unfortunately finding your way in documentation is not always easy. Without any help, it is a real challenge to find documents that are actually helpful among the many on the MirrorLink page or the Car Connectivity Consortium site.
Do not let that discourage you! We have already fought our way through the tangle of documents and will gladly share our knowledge with other software providers. To help you easily adapt your app for MirrorLink, we have compiled instructions with the most important information.
Once MapTrip was running in the car with MirrorLink, the results of our tests were genuinely pleasant: the app is presented true to the original with no change in its layout and is easy to operate.
This seems like a perfect smartphone-car environment. Unfortunately however, the MirrorLink inventors came up with the idea of app certification. Before a MirrorLink app is published, it has to be reviewed by an authorised body. Certification reviews whether the app is “car friendly”: Are the contrasts adequate, can the app be used without distracting the driver, are the fonts large enough and legible etc.?
Several certification levels are offered. Level one allows the app to be used while the car is parked. The purpose of this usage scenario is unclear to us. At the next level, the app may also be used while driving. Additional levels verify the app for use in the USA or Japan. A four-figure certification fee in Euros is payable in each case.
In the meantime, our MapTrip navigation app works beautifully and also looks really great in the car according to our own tests. Certification is now pending and we look forward to the result with anticipation: some fonts in MapTrip may be a bit smaller than specified at certain resolutions. Adapting these would require a fair amount of effort. This could lead to a slew of changes. So we decided to just wait and see, and we are curious about the result!
The major advantage of the mySPIN technology from Bosch is that it connects both iOS and Android apps to the car! The application is mirrored on the screen in the vehicle in a way that makes sense to the driver. The app can use the car’s GPS and speakers.
A mySPIN app also receives signals, for example when the steering wheel buttons in the car are pressed. New in-car app control can therefore be developed which is particularly deflective and offers real added value against the direct use on the smartphone. The only other thing we could ask for is more vehicle data, for example how much is left in the tank or the remaining range of the vehicle.
Unlike MirrorLink, mySPIN does not transfer the screen contents of the smartphone to the car 1:1 but optimizes the layout suitable for cars. App dialogues are displayed in the landscape format, which means the layout is in part adapted to the vehicle display by the system. Therefore the developer may have to make changes.
Similar to MirrorLink, there are comprehensive layout rules for mySPIN regarding contrasts, button sizes and the like. These have to be followed. MySPIN is developed by Bosch SoftTec. That is where you will find information and documents, and where compliance with the design rules is ultimately verified as well.
To use mySPIN apps in the vehicle you first need what is called a launcher app. It establishes the connection to the vehicle and therefore has to be installed on the smartphone in addition to the actual mySPIN app. The launcher app is provided by the OEM. In order to use mySPIN apps in our Land Rover Discovery, we therefore had to install the “Land Rover InControl Apps” app first. A comparable app is also available from Jaguar, which uses mySPIN as well. OEMs also use these launcher apps to control which mySPIN apps they approve for use in the vehicle. A mySPIN app is therefore not automatically usable in the car but must first be released by the respective OEM. An important difference e.g. to MirrorLink!
The mySPIN technology actually supports both Android and iOS. While this is fundamentally a big advantage, both an Android and iOS version is available of just about any app that wants to be taken seriously. Like MirrorLink, the current mySPIN version needs a cable connection between the smartphone and car. However, this may be replaced by a WLAN connection in a future version.
Bosch SoftTec has recently offered a solution for using mySPIN-enabled apps in vehicles that are equipped with MirrorLink. This works with a hardware adapter, which must be specified by the OEM. Therefore, drivers not only have access to the extensive mySPIN app portfolio, but iPhones, to which MirrorLink is not compatible, can also be connected and used in the vehicle.
mySPIN is as mySPIN for 2-Wheeler recently available for motorcycles. Here mySPIN for 2-Wheeler helps the developer tremendously in achieving good usability on the bike. After all, the smartphone naturally cannot be operated with heavy gloves while riding. If the motorcycle has integrated buttons, these can also be used to control the app. Bosch is heading in a good direction with mySPIN for 2-Wheeler since the range of app functions and usability in general are dramatically improved.
We demonstrated this with our “killer curves” app. The application warns motorcyclists of especially dangerous curves with an alert on the motorcycle display. A corresponding app runs on the smartphone and monitors the ride. When approaching a killer curve, a corresponding warning can be issued through mySPIN for 2-Wheeler. This makes smartphone-car integration fun – or actually smartphone-bike integration!
Further information to mySPIN:
Anyone dealing with Smart Device Link (SDL) will soon stumble across certain terms, so these are briefly explained here: Smart Device Link is open-source software to connect the smartphone and car. The development of SDL is mainly driven by Ford and its subsidiary Livio. The Ford-specific version of this technology is called AppLink. In the car itself, AppLink apps can only be used with a SYNC infotainment system.
SDL is currently found only in Ford vehicles. Toyota is implementing this technology as well but has no corresponding vehicles on the market yet. SDL is compatible with both Android and iOS devices. iOS devices have to be connected with a cable (for now), but for Android the connection already works with Bluetooth.
Unlike MirrorLink and MySpin, SDL does not transfer the screen contents from the smartphone to the vehicle display. Instead the SDL apps display very simple dialogues in the car. The layout of lists and other GUI elements is basic and there is little or no room for individual design. Graphics can only be integrated in a very simple form. What may sound negative at first also has concrete positive aspects: an SDL app can be implemented in next to no time since there are no elaborate layout and design considerations. The app-specific layout however is definitely lost. Some proud developers may find this very objectionable.
We were fortunate to be provided with a simulator for our work with AppLink. This “Ford in box” contains the SYNC infotainment system and a number of buttons and switches that can be used to simulate the rest of the car. One button can be used to turn on the ignition, another to start the engine, and a third for instance to turn on the headlights. All this right at your desk!
Ford provides very good documentation. This includes lots of sample code, instructions, and everything else to make a developer’s life easy and productive. Ford consistently works with developers here similar to Google and Apple.
In a coming version, AppLink is slated to also support the transfer of screen contents to the vehicle display. This can be used for instance in a navigation app to display the map. Various approaches are being mixed here again, which is also found with MirrorLink and others.
More information is found at:
FORD Developer Program
OpenCar is the newest member of this squad and a blatant outsider for several reasons. OpenCar was developed by Jeff Payne who founded the company. He obtained venture capital and locked himself in with 25 developers for three years so that, as he told us, he could work undisturbed “in stealth mode”. During this time Payne and his team developed a highly comprehensive system that, in addition to apps, is for example also able to control a car’s comfort features.
Unlike all of the previously described systems, OpenCar does not need a smartphone at all. OpenCar is not about somehow “transferring” a smartphone app to the vehicle. Instead OpenCar provides an environment for running special apps. This means OpenCar apps do not exist yet, they have to be newly developed!
OpenCar was acquired by INRIX in 2016. Until then INRIX was primarily known as a provider of traffic and parking information. Today OpenCar is probably INRIX’s most important product – and is being expanded into a comprehensive software environment.
infoware was in Las Vegas as a participant at the 2nd OpenCar Coalition Workshop. There we presented the first OpenCar app and shared our experiences with the attending participants from around 20 OEMs. We saw that most of the participants were collecting initial information about OpenCar.
Since OpenCar not only offers a new technology but also an entirely different concept and business model, there is much that is new here. The idea behind OpenCar is to offer a platform for software providers and OEMs that meets the needs of both sides. Software providers quasi develop only the skeleton of an app. Then the app is published in the app store in this generic form. OEMs select the app in the store and then adapt it to their own design themselves.
From the perspective of software providers, this is an important prerequisite to make the car market attractive. After all, adapting one’s own app to the different design specifications of Daimler, Renault or Volvo is generally a very complex undertaking. On the other side, the OEMs can choose from a (hopefully) large number of apps and adapt them to their brand design with little effort.
In our case we were even able to use existing code from an HTML5 navigation prototype.
To date our practical experience with OpenCar is limited to developing and testing our app in the simulator. The documentation, support and simulator make a very good impression. Building on our own navigation API, we made progress surprisingly quickly and put together a simple navigation app within two weeks.
To be fair, it needs to be said that problems often begin with other technologies only when transitioning from the simulator to the vehicle. We were not able to collect any experience here since cars with OpenCar do not exist yet.
OpenCar is still in its infancy and we were surely in a privileged position due to our close contact with INRIX. Support was certainly influenced by the fact that we as a first mover wanted to present our app at the OpenCar Coalition in Las Vegas.
That being said, we really like what we have seen so far! The documentation is complete and well organised, the simulator works well with no major hiccups.
More information about OpenCar is found at:
We look forward to seeing how the technology of the various players continues to develop and who will be the most successful competitor for Google and Apple in the car going forward. Our work will continue!