The Soft Car

 

Gil MŁller

 

15.11.06

 

This paper sketches on outlook for software services in the car. It considers a car not only as a transport vehicle but also as a mobile personal space similar to a home. This view introduces a tension between the safety and efficiency requirements of transport versus openness and flexibility of IT services. This means the architecture for such a car has to be a hybrid with characteristics of real-time systems for the carís core and elements of desktop systems for its shell.

 

 

Outlook

 

A car is mainly designed and optimized for transport. Additionally, it shares functionality of a home. It is a mobile personal space with infrastructure, which provides comfort, shelter, storage facilities and last not least communication and infotainment facilities. Briefly, we will divide the car into a transport facet (the carís core) and a home facet (the carís shell).

 

Particularly, the home facet is of interest for us: there the world of mobile computing hits the automotive world.The recent tide of mobile computing in terms of devices and services impacts the design of in-car environments. Car nowadays support digital multimedia services (like CD play-back, navigation and MP3) and mobile digital communication. Nevertheless, the introduction of such facilities in the car is much slower due to the long design time and life-cycle of cars. This leads to tension with the more rapid progress in the IT industry.

 

So far, the carís digital infra-structure is rather closed and dedicated. It links only the in-car devices and is not meant to be extendable Ė for the end-user. For good reasons: to ensure the safety of the driver and the passengers and to guarantee a unified design. Still, users take mobile devices into the car and will want to use the carís facilities with them. A brief discussion of the current four main uses for mobile computing demonstrates that:

-          Telephony: This is by far the main use of mobile computing. It is now usually not directly provided in the car. Instead the car provides a framework which enables a mobile phone to enrich its service. In terms of industry this is a mature product. But still support for it is not running smoothly as it should.

-          Infotainment: This is a wide field including audio, video, audio broadcasting, TV, games, and last but not least the web. So far, the car supports traditional media (e.g. CD playback or analogue audio broadcasting. But, this use case is also one of the hottest (with the most rapid path in the IT industry) and of broadest interest. It is unclear, how the current carís IT architecture can follow this rapid development.

-          Navigation: This used to be a domain of the car. But, with the advent of PDAs and map services on the internet, its usage is now far more broader. It could be kept as a dedicated service in the car. But, it should be offered with more linking capabilities: e.g. integration with PIM, connection to internet services, and support of smartphone/PDA navigation.

-          PIM/Collaboration Support: This includes not only classical time planning and contact management, but also (push) e-mail. In future, we could even expect more dedicated service with more networking (mobile office and logistics). This requires a strong integration with other mobile devices.

 

In all cases, there exist devices which are either dedicated to the mentioned use cases or even support all of them. Where is now the role of the car? Its strength is its unified and solid design. Additionally, it could provide a service quality which reaches that of the home (e.g. it can offer video and audio devices which are much better than those of mobile devices). But, our analysis also showed that there is a strong need for a more open and extensible in-car environment.

 

Architectural Considerations

 

In general the carís infra-structure shall be extendable, open, sustains the carís life-cycle, and holds the carís service qualities. This matters the end-users, the car maker, third party equipment makers and service and maintenance providers.

 

The following alternatives exist for the car IT:

- Car IT has its own path (the current situation), where the industry has problems of catching up with the IT industry, or meeting the expectations of customers

- Car IT migrates to it industry standards (Windows, Linux, Ö)

- A gateway is defined by industry as by OSGI

- A hybrid strategy is pursued (e.g. automotive standards for internal components;IT standards for the service environment)

 

A hybrid strategy is the most plausible alternative. The carís core with its high needs in terms of safety requires a rigid, closed and mature design with reduced complexity. This would also enable to use of best practise industry standards. The carís shell needs to be based more on the IT industry standards in order to be compliant with state-of-the-art services and devices.

It should be divided into a guarded and into an unguarded area. The guarded area would provide applications and services, unlimited access to the in-car environment. In order to ensure a unified design and safety, such services and applications would have to be certified or at least be controlled by the carís OS (similar to the sandbox model of Java). This could be done on the basis of a gateway. For services and applications in the unguarded area strong restrictions apply with respect to the carís resources. This would mean that such entities only provide data or are performed in a secured area (e.g. using virtualization techniques like applets, browsers, or virtual machines).

 

This design provides for car operations where change is not needed during the carís lifecycle. The guarded area serves for controlled extensions and the unguarded area for rapid changes.

 

Business Implications

 

This design would let the car industry meet the customer expectation with moderate effort and risk. Additionally, it provides an evolutionary path for its current IT strategy. Particularly, the guarded area is a means for new business for the car industry:in terms of IPRs, licensing and certification to third parties, and offering new services and applications in the after-sales market. This strategy leads to a soft car: a car that morphs itself to a gestalt in terms of devices and services suiting the userís demands.

 

© 2006 Gil Müller (www.gil-mueller.com))