For more than 25 years experimental problem solving has been an integrated part in many of the science and engineering programs at Linköping University. Our main implementation of the experimental problem solving is done via a type of laboratory work where the students have access to an experimental setup with the task to obtain an equation that describes the experiment. It could, for instance, be the task to describe the period time for some type of pendulum or the sinking time for an object in a water tank.

The procedure to obtain the descriptive equation include; hypothesis (what variables should be included in the equation and how should they be included? Is the equation a product or a sum etc.?), dimensional analysis (to obtain exponents), measurements, measurement data handling (logarithmic plotting etc.) and error analysis. In this way the students will practise skills which are important to master as a future researcher or engineer. This type of laboratory work has been included in the introductory courses in several of our engineering programs (e.g. electrical, computer, mechanical and chemical engineering programs). In Linköping, the concept has also been used as cases in problem based learning (PBL) courses.

Recently, we have developed a LabView™-based software application (VExLab) for “experimental” problem solving. The main difference is that the measurements are now done virtually in a software “simulating” an experiment. We are careful to stress the importance of real experimental work in engineering programs. However, as a complement VExLab is interesting for several reasons. For instance, the software approach makes it possible to; • distribute the problem solving over a larger time-span. The students are not limited to given scheduled occasions but can solve the problem at any time at the university computer facilities or on their own computers at home. • construct problems that are too expensive and/or difficult to do experimentally in a teaching laboratory. • make individual problems for each student group, e.g. in terms of introduced errors. • construct problems focusing on some special part of the problem solving process. For instance, the dimensional analysis or error estimation. • construct new problems on a much shorter time scale. • construct problems connected to a broader number of research fields. • construct a number of mini-problems connected to a specific course. The use of LabView™ also makes it possible to include signals from real detectors in the simulation.

The experimental problem solving will be demonstrated in collaboration with the audience. Both the experimental and software concepts will then be presented and compared. We will also describe how the two approaches is typically implemented and combined in our courses including the examination through written reports.