Editor | On 22, Jan 2018

Darrell Mann

Constructing Function and Attribute Analysis (FAA) models is one of my least favourite TRIZ-related activities. At the same time as trying to avoid drawing them, I also know it’s the most important problem definition job I have to do. A frequent short-cut involves trying to draw a minimal model featuring as few components as possible. This is good for focusing in on a problem symptom, but not so good from the perspective of avoiding unwanted side-effects when we fix a problem in one part of a system and unwittingly create a new problem in another part that we had failed to model.

A good heuristic to keep in mind when deciding where to draw the boundaries of your next FAA model involves making use of the Law of System Completeness, and specifically, making sure that all of the six essential elements are present for the main useful function you’re modelling.

To take a simple example, let’s say we’re looking to construct a model of a windscreen wiper system for an automobile. The main function of this system is clearing the screen from whatever environmental contaminants might be present. A typical system model might then look like the one illustrated in Figure 1:

Figure 1: Exemplar Function Analysis Model Showing LoSC Elements

Clearly, there are more components we could have chosen to add to this model, but we’ve drawn the boundary where we have because we can see that the elements that allow us to deliver the ‘clear screen’ function are all present. The screen represents the Interface; the thing that does the clearing job is the blade (‘Tool’); the Tool is powered by a motor (‘Engine’); the Engine and Tool are connected by a cluster of (‘Transmission’) components – motor spindle, blade-arm and frame – and the whole system is Coordinated by the driver, who Senses when there is a need to switch the wiper system on.

In some vehicles, the driver might have been replaced by a rain-sensor and a software algorithm that tells the motor when to switch on, but the Law Of System Completeness nevertheless continues to apply. And as such our model would be well advised to feature all six of the essential elements.

That said, having the six elements present doesn’t necessarily imply a need for six components. In the Figure 1 example, the driver fulfils both the coordination and the sensor functions. The sensor ‘component’ in this case has in effect been subsumed into the complex system that is the human operator.

As the windscreen wiper system evolves towards its ultimate Ideal Final Result, it is the case that progressively more of the components of the original system will disappear. In the ultimate situation, we will likely end up with the FAA model illustrated in Figure 2:

Figure 2:  IFR Windscreen Wiper System

Although this hypothetical Ideal Final Result screen now only possesses one ‘component’, crucially, the six elements required by the Law of System Completeness must necessarily still be present. The trick is that they will have shifted to the sub-system level.

Let’s imagine, for example, that the self-clearing screen becomes possible by adopting something like the Lotus Effect shown in Figure 3.

Figure 3: Lotus Effect

If we were to now construct a FAA model of this system, it might look something like the image illustrated in Figure 4. The model has been constructed by using the Law Of System Completeness as a checklist: most visibly, the Interface in this new system has now become the rain and the Tool is provided by the 3D protrusions on the surface of the glass. Finding the ‘Engine’ is a little more difficult, but we know it must be there because ‘something’ has to provide the ‘field’ (source of energy) that our minimum system must contain. A little knowledge of the Lotus Effect will tell us that our ‘field’ in this case is all about surface energy, which in turn is about judicious design of the 3D protrusions to make van der Waals forces work to give us a surface tension effect on the water droplets so that they don’t wet the surface.

So far so good. Next up comes the ‘Transmission’. This is the ‘thing’ that will connect the Engine to the Tool. In this context, it is the underlying structure provided by the glass sheet that is doing this job.

This leaves us with apparently missing Coordination and Sensor elements. These are clearly rather more subtle. The Coordinator in the LoSC is the ‘thing’ that controls the delivery of the useful function. There is no specific component doing this job in a Lotus Effect windscreen, but rather the useful function is delivered ‘automatically’ by orienting the glass at an angle that ensures the rain runs off the surface. Which in turn tells us that part of the ‘Engine’ must also be gravity.

Finally, we’re left with ‘Sensor’. Here, at the IFR solution, we see that the ‘need for measurement’ is eliminated because the system inherently delivers the desired behavior. In this sense, the Coordination function of the glass incorporates the sensor: as soon as a water droplet arrives, the glass ‘knows’ and the van der Waals forces do their thing.

Figure 4: LoSC-Based FAA Model For Self-Clearing Screen

Figure 4 now offers the basis for a continuation of the evolution of the screen. We know, for example, that the Lotus Effect by itself probably is insufficiently effective (in its current form at least) to deliver some of the other functions of the windscreen. Not least of which would be the ‘high-speed insect impact’ problem and the need to keep the screen clear of organic matter. At which point we realise that, if we’re to successfully also achieve this function, we must also satisfy the Law of System Completeness. And that takes us full circle back to the first article this month… can we get one complete system to deliver multiple functions?

Like, for example, switching to autonomous vehicle designs that don’t require the driver to see through a screen anymore. Or at least not using a sheet of clever glass.