ZSPACE Revs Up New Showroom AR Platform for Ferrari

The customer begins by taking a look at the cars displayed, choosing a model to find out more about - and jumping into the app. It approaches each car’s characteristics in seven categories, accessed as interactive video, text and image ‘activations’ that show off particular aspects of the aerodynamics, braking system, engine features and powertrain, aluminum chassis and car customisation.

Ferrari Interactive

Using the camera in the iPad, the customer stands in front of the car and holds up the device to position the car in the centre of the screen. A graphic outline, or watermark, of the car appears for the customer to line up with the car onscreen. As soon as the two are aligned and the watermark locks onto the car, interaction begins.

Customers tap various options to access information in interesting ways – like watching the engine or wheel construction dissemble and reassemble before their eyes in an elegant animation that reveals the important points for drivers. Or they can read a slide-out text window about their model’s braking system, or see how air flows along the aerodynamic body design.

Executive producersTim JardineandSam Hallat ZSPACE talked to us about developing the project, and working with Ferrari to make sure it proceeded according to what they needed. “This app developed entirely out of an idea initiated by the client, instead of Ferrari approaching us and having to fit their requirements to an existing implementation at ZSPACE,” Tim said.

“For example, the topics and material for the seven activations came from the seven-step introduction in a brochure Ferrari had designed earlier on.  They were very receptive to the newer, digital approaches we suggested, and were especially keen to incorporate interactivity as a key attraction for prospective buyers. They are recognising a new kind of Ferrari owner that up until now may have felt slightly intimidated in a Ferrari showroom. Now, they have a personalised introduction.”

ZSPACE’s developers and 3D team spent about 18 months all together on this project – three months refining the concept and outcomes, and another 12 months on R&D and production. The work began with the cars. The 3D team built full, extremely accurate 3D models in Maya of each Ferrari model to its exact specifications based on thousands of images and hundreds of hours of video.

Model Tracking and Recognition

A very important use of the modelling data was creating and calibrating thewatermarkfor each Ferrari model by identifying an optimum number of clear, distincttracking pointson the body. The watermarks were a critical step because they are the user’s way of telling the app which model he or she is seeking information about. As soon as the watermark locks onto the car’s tracking points and ‘recognises’ the model, the correct app launches.

The 3D models, or portions of them, also appear within many of the activations – not only the exterior but also internal parts, calling for accurate modelling and animations. Look, design and authenticity were key factors. Ferrari’s F1 heritage is something the company likes their customers to feel involved with. For example, their racing brake and wheel systems have been partly applied to the road models, and so the app gives the customer an animated, insider’s view of how they are built.

The development team usedMetaio, an application with specialised tools for augmented reality application development and markerless tracking, to lay the app’s groundwork. Because they were working with programmers, they chose the software’s SDK instead of a prepared framework version. This gave them more opportunity to refine the details of their app, but also meant a lot of iteration to control the results, especially since Metaio was still in beta when they started the project. 

On the Horizon

Because tracking is such a critical element in the app’s effectiveness,environmental lightmakes a difference – full, even light works best, creating a low-contrast, low-reflection scene. However, ZSPACE would like to work on improving the app’s vision enough to make the app usable outdoors and under varied lighting conditions.

Apart from learning about the cars, customers can use the app’sConfiguratortool to customise the Ferrari to their exact specifications and superimpose the paint work, styling and trim options as CG elements onto the showroom car. As they compare and make choices, they can record short 15-second videos, save these out and email them to their friends and themselves to look at later at home.

The Ferrari AR Showroom App is built as an upgradable, customisable platform as well as a finished application. The panel text can be updated and edited through itscloud-based CMS, and Ferrari will be supplying local-language translations for their Japanese showrooms in this way. Also, while the app currently showcases five models, the company can add or retire models and update options and other content in real time through the CMS. Thelive analytics, currently used to monitor how customers use the app, could be extended to collect other data.

Sam Hall said, “One of the greatest potentials of this Showroom AR App is that, with minimal R&D, it could become an offsite application. We have tested the app outside and in the Ferrari workshop and found that, because it is based on real car contours and tracking markers, it could be used anywhere a car can go.” www.zspace.com.au