Frank Weber is GM's straight-shooting vehicle line executive of the Volt program. I had the chance to ask him what was happening with mule development.  At this point there are 35 or so developmental mule vehicles.  Mules are early prototypes that have the full Voltec drivetrain but a borrowed interior and exterior.  Within days the engineering freeze on the first true Volt prototypes or integration cars will occur.
What is the current status of the mules, since you are done with this phase, and where are they in the testing process?
Actually what is happening is we have done all the testing as planned last year. We have been using the winter for winter tests. You are counting in summers and winters. This is a very important phase to test everything under very cold conditions and this is what has been happening now.

So are you still figuring out the control algorithms and computer code?
This is all the standard procedures in a car and you know what to do and we have all those controls incorporated into the cars. But now what's happening is the true development work that you say OK this is the temperature of the battery, and this is the temperature of the system, and this is what happens when you are plugged in, etc.

There are parameters that we call calibration, you have the basic software functionality on those cars defined, and then we start to calibrate it looking at the temperature and when to we start it, what is the true power of the battery at a certain temperature , etc.

How do you know how the batteries are performing in the mules, and how can you extrapolate forward to know that you hit the right sweet spot to make the battery last 10 years?
What you know is what the behavior is for the cars that we are testing, and then you make an assumption for how a component will behave over time and how it will behave under the same situation in several years.  This is what we call accelerated testing. This gives you some indication of durability.

The piece that is tricky and interesting about the battery is to do a really accurate extrapolation of the true behavior. For a mechanical part this is very simple. For a mechanical part you can replicate its lifetime and find out when it will break.

The battery is electrochemical and its more difficult to make those extrapolations. This is part of the learning we have to do, battery learning between the battery supplier LG and us. By the way this is still the element of risk. This is also why we are unable to get the car out any sooner. It is those things that have to be developed now with the components that are representative of the production vehicle.  There is no way to do this any faster.

Have you started to build the integration vehicles yet?
No this is a different phase. Mule cars are now completed and have been since last year. They were all built on time. Whats happening now is those mule cars are now being tested up until the middle of 2009 and then they are replaced by the next generation of vehicles which are called integration vehicles. It takes a couple of weeks to complete the build.

Are the internal parts on those more refined than the current generation.
Yes, always. Whats true in the development is if you find something you might make a small adjustment to it to improve it. Its an updated version of those components.

Are you going to take those integration vehicles out of the test grounds and drive them around the real world?
Normally you have to be a little bit careful, but since we've shown the world the production version we can in principle take them wherever we want.

Will you?
Only if it is value added. There are a couple of interesting streets in the world where development insight is really generated. Then it might make sense to take then out there. By the way the current mule cars are being taken out on the public roads.