Apps 2 Automotive will be held again in July 2011!

Automotive & Spice Days 2010

tuesday, the 21st of June and on wednesday I took part in the A2A symposium which was part of this years Automotive & SPICE Days.

After the opening address by Dr. Walter Rogg from the „Wirtschaftsförderung Region Stuttgart“,

the first keynote speaker was Tom DeMarco. He gave a talk on „The mojo of systems“. Very inspiring!

He explained to us, why our brains are to small to understand how systems work and that we carry more information in our pocket (on a USB Stick or Smartphone) then our brain could ever possibly hold. We learned that geeks are not antisocial, giving trust creates a link and people who give trust are leaders.

The other talks I listened to were as follows (in German):

• Mobile Betriebssysteme für Smartphones im Vergleich Wolfram Herzog, SIC! Software GmbH
• Was heutige Fahrerinfo- und Assistenzsysteme zum Nachrüsten leisten. Die Kluft zwischen Kundenwunsch und Umsetzung, Hans-Peter Schmidt, Krämer Automotive Software
• Bedienkonzepte der Zukunft – Das mitdenkende und mitlernende Fahrzeug & Mobile Endgeräte im Fahrzeug Richard Mutschler, GIGATRONIK Stuttgart GmbH
• Eine Methode zur Bestimmung der optimalen Rückmeldemodalität für Fahrerassistenzsysteme Lars Lewandowitz, Karlsruher Institut für Technologie
• Links abbiegen, dann links abbiegen: Interaktionsprobleme heutiger Navigationssysteme Florian Schaub, Universität Ulm
• Bedienkonzept Audi: Die nächste Generation Dr. Werner Hamberger, AUDI AG
• Modellgetriebener Ansatz zur flexiblen Integration neuer Applikationen ins Fahrzeug-HMI Mark Poguntke, Daimler AG
• Car-to-car Kommunikation: Funktions-Enabler oder Privacy-Alptraum? Dr. Elmar Schoch, Universität Ulm
• Warum LTE die “Automobile Kommunikation” verändern wird Ulrich Möhlmann, Alcatel-Lucent Deutschland GmbH

They all were very interesting, some details follow below.

At the evening dinner party we had an also quite famous guest speaker: Gunter Dueck an IBM Chief Technologist!

He gave a very funny talk on how to create a smarter planet and IBMs cloud computing strategy: „it is very simple, all we have to do is rewrite every software! No problem, we did the same, when the washing machine was invented, didn‘t we? We bought all clothes new!“

On Wednesday the keynote speaker was also very famous: Prof. Bjarne Stroustrup, the developer of C++. He gave a rather technical talk on safety, performance and productivity with C++.

After his talk I had the chance to ask him a question: „What do you think of new languages such as, e.g. D?“ (D – the so called successor of C++). His answer was rather unexpected: „What language? D?“. He did not seem to know D. Well, he probably does not need to know it…he is the C++ guru!

The mojo of systems, Tom DeMarco

Tom asks the question: „What is a system?“

He says: „We know a lot about systems, but we can‘t explain,

what a system is.“

How would Aristoteles define the term „system“?

Definition of the definition (by Aristoteles):

Definition = genus + differentia

For example: Man is the animal possessing the capacity to
articulate speech.

What is a System then?

A System is a complex interconnected set of components
that act together to achieve some end.

Components themselves may be systems.

Systems are hard to understand: Who understands Windows or Photoshop?

Why are systems so hard to understand? Answer: The human brain memory can not  hold it.

Tom tries to explain this supposition:

In the information theory code is a message: For example, a project to build a software system is a message to a computer.

Scary observation: Every bit that is in any of the messages that make the system has to have been inside a human brain at one time or another.

Tom explains to us, how much memory a human brain can hold:

He cites John von Neuman, who had some good news for us. He said that a Human brain can hold 10^20 bits in memory.

But, Thomas Landauer said von Neuman is wrong.

Landauer stated that every information to our brain comes through our senses. The sense with the highest bandwidth is our sight.

What is the bandwidth of the eye? Answer: 10 MBit/s.

If you do the math, this would mean that when taking into account that you do sleep 8 hours a day, when you are fifty you would have seen something like one billion seconds: one billion seconds times 10 MBit/s = 10^16 bits

Bad news:

Landauer said that the human brain only absorbs 2 bits per second: the brain memory can hold 2 Gbits.

A 2 Gbit Memory stick is 2 $! Our brain is worth 2$!?

But our mind is multi-dimensional.

Well then, how do we deal with complex systems at all?

• In the old days we used to absorb the systems completely, this is not possible anymore.

• Today we divide systems into small pieces loosely joined.

But: natural systems are composed of large pieces tightly joined! That is why it is nearly impossible for us to grasp natural systems such as the weather system.

The optimal (man made) system is: smallest pieces most loosely connected!

This means the amount of collaboration is increasing dramatically.

But how much collaboration do we need?

Answer: Total project collaboration >= message size of the project

That means, that geeks are doing more collaboration than has ever been done before: Geeks are not antisocial!

Most important in a project are collaboration and design. Conceptional design and internal design.

Conceptional design can be done by a small team but the internal design has to be understood by everyone.

Today, we are constantly raising the bar for how much collaboration has to be done.

Tom shows a picture of the book „The design of design“, by Frederick Brooks.

But there is another important point:

If I receive an email from Nigeria, from someone writing to me, that he wants to give me a lot of money, I get one bit of communication: Delete! But why? I do not trust the sender! What does this mean?

Karl Erik Sveiby says: Trust is the bandwidth of communication!

Trust is very important for good collaboration. Things we do to develop trust do not translate through the telephone. We need to see the people.

He gives an example:

When your young son comes to you saying: „Dad, I have done it! I passed the driving test! Can I have the car tonight?“ You would swallow, but give him the car key, wouldn‘t you?

Why? Because he demonstrated trustworthiness.
But, giving trust in advance is necessary! Giving trust creates a link: People who give trust are leaders. You give trust to get trust.

Giving and getting trust happens over lunch not in a work meeting.

Recap, or what we should remember:

– Aristoteles definition of definition: Definition = genus+ differentia

– Your system may be bigger than you

– The brains bandwidth is 2bps

– Small pieces loosely joined make a good system

– Collaboration and Design (internal design)– two most important skills

– Trust is the bandwidth of communication

– Advance trust before demonstrated trustworthiness

Session Talks Intro

Dr. Tscheschlok (Gigatronik) says a few introducing words.

He mentions ask a few interesting questions:

• Will we have the iPad in the car at Audi?
• Will we have an infotainment client-server architecture?
• Ford: DriveSmart – cloud based services in the car – the future? Ford uses Windows Embedded
• Conti: – melting of domains – mashups in the IVI
• Always on in the car: Emails while driving. Multi-media high-end functionality in the vehicle
• Seamless handover between handheld and car navigation systems
• Which new hardware platforms will be required to keep up with the software
• What to do about the product cycles: Car – IT discrepancy?
• Challenge: brand and market specific requirements

In the following a few of the talks are discussed:

Mobile Betriebssysteme für Smartphones im Vergleich (Mobile operating systems compared),
Wolfram Herzog, SIC! Software GmbH

Mr. Herzog introduces his company at first and then talks about their study concerning comparison of the following different mobile operating systems/platforms:

• Symbian
• JavaME
• BlackBerry
• Windows Mobile
• iPhone OS / iOS
• Android

There are many more platforms on the market but these are the most important ones.

Each platform was tested on the following criteria:

• Technical feasibility – what is possible to do with the platform
• Development effort
• Supported functionality
• Maintainability

Technical feasibility

He shows matrix of functions derived from 12 software projects.

Result: a deep integration is only supported by Symbian and Windows mobile

Development effort and supported functionality

The highest development effort is required on the Symbian platform in native C++. The interesting fact is, that when using the new QT based development IDE for Symbian the effort is nearly equal to iPhone or Android development. So QT makes Symbian development much easier.

The major problem with iPhone development were the many restrictions.

Maintainability & Testing

• Symbian and Blackberry are quite difficult to maintain. Symbian QT improves dramatically
• Highest testing effort: MIDP Java due to the different platform support. High testing effort also with Symbian and Windows Mobile due to the many different platform versions and inconsistencies.
• Testing effort increased dramatically with Android due to the different OS versions

Who wins, who loses?

• MIDP Java, rather dead. No solutions for new usability requirements. Signing problematic. Bad standards: reserved parameters are interpreted differently by the OEMs. No touch concept
• Symbian OS: Native Symbian OS not maintainable,  QT technology as future basis leads to massive recoding of all Symbian apis and gui routines. Future smartphones shall have a common OS basis. Goal: lifetime over the air and don‘t underestimate the Finns!
• Android: is in danger because of too many cooks. OEMs have to differentiate somehow this is currently done by the UI concept and the Android version. Danger of fragmentation exists
• Blackberry: Where are the customers?Problem of closed company installations. No free internet, no market for apps

When will the iPhone killer arrive?

When we look at the past developments in IT history nearly all 11 years a new technology arises. PCs in the 80‘s, Mobile Phones in the 90‘s, iPhone in 2007. This may lead to the suggestion that 11 years after 2007 (year of the iPhone) the next generation of smart devices will arrive: 2018! We are looking forward to it!

Was heutige Fahrerinfo- und Assistenzsysteme zum Nachrüsten leisten. Die Kluft zwischen Kundenwunsch und Umsetzung. (What contemporary exchangeable IVIs provide. The gap between customer requirement and realisation). Hans-Peter Schmidt, Krämer Automotive Software.

Mr. Krämer starts with talking about different studies they conducted on customer groups requirements. What are the requirements of elderly people, what do younger people expect? Each of the customer groups have different driving abilities, problems and habits. This has to be taken into consideration when implementing a navigation system.

He asks the question: why are there so many iPhone apps available but nearly none in the car?

One of the main reasons are the very different development cycles between the automobile industrie and the IT. The only solution to this is to provide two open DIN slots including a standardised and secure connection to the car electronics.

A problem with this solution is that the OEM fear of losing a market (selling expensive built-in IVIs). But the reality is that the market already is lost. They should be open for a new market where the OEMs provide the exchangeable IVIs themselves.

Bedienkonzepte der Zukunft – Das mitdenkende und mitlernende Fahrzeug & Mobile Endgeräte im Fahrzeug. (Future concepts of operation – the thinking and learning vehicle & mobile devices in the vehicle). Richard Mutschler, GIGATRONIK Stuttgart GmbH.

One of the main questions is: How can we assure safety and high comfort when using electronic devices in the vehicle?

The main problem is driver distraction. How can we assure that the driver is not distracted by the navigation system or mobile device?

Different sensors and alternative modes of operation as well as new usability concepts are required.

Tests have shown that most users prefer one mode of operation, e.g. speech control or touch, when communicating with a device. If one modality is not available (e.g. speech, because of an incoming call) a switch between modalities must be supported by the system.

Future speech control systems will allow a more kind of dialogue between the driver and the system.

Another problem is the integration of mobile devices as there are too many different devices and no standards. Three different approaches currently exist:

• Device is primarily used as data source (MP3, Videos) and is controlled by the built-in system
• Terminal mode: the display of the device (its user interface) is transmitted to the built-in system
• Remote application: parts of the built-in user interface directly interact with the applications on the mobile device

A goal of the mobile device integration is the uninterrupted data and communication flow. This requires new standards for device integration (hard- and software) and the apps need to be safe in the way that they do not distract the driver. A „safe mode“ of operation needs to be standardised in the car to mobile device interface.

Mr. Mutschler shows their tools to develop new HMI models and implementing them using the Model Driven Architecture (MDA) approach. The HMI tools are used for simulation, concept scetch, generating target code, developing a testmodel and defining OEM specifications.

Most important success factors for mobile device integration in a vehicle are:

• Standardised interfaces to the vehicle
• Domain independent harmonisation of user interfaces
• Application of user centric development tools

Liability is one of the concerns that need to be clarified when integrating mobile device in the vehicle. Who is made  responsible is the driver is distracted and has an accident?

For further information on the A2A symposium please do not hesitate to contact me.