Anticipation, Windshield Wiper, Part 1

I presented an example of a windshield wiper as an easy exercise for anticipation (Anticipation Part 1, Part 2, Part 3, Part 4). The following checklist helps to understand the model of anticipation. The “real” checklist involves a lot more elements, but I have cut and rearranged some elements for the sake of comprehension. Furthermore, it cannot depict the dynamics. I will present other analysis tools for these considerations (see also the article of Wildcat Hendricks). So, at the moment, this list is just a start. It indicates if the attempt to anticipate a situation fails. Usually, all these elements are necessary for successfully using anticipation. Some things on that list look redundant. Still, these points are necessary to broaden the perspective. It is important to analyze the emerging problems first. There are numerous work techniques to overcome determined problems. So, don’t hurry and jump to rash conclusions. It is possible that you quickly find other solutions. But be careful, these might not be based on anticipation.

Please recall the windshield wiper example. The outcomes were already discussed in the blog entry Anticipation, Part 4. Now, as an apprentice piece, we use the checklist. The second part of the example with variable speed and displacement is more interesting.

 

EN_Scheibenwischer_02

 

1. Are patterns or traits available? All the time?

You don’t really think that your partner will shake his hand in advance to announce the upcoming directional change, do you? Do you remember the example of Scissor, Stone, Paper? It is not possible to differentiate between the alternatives if the needed patterns and traits arrive too late. Thus, you have two separate problems. The upcoming change offers no advance information and even if it did, it is still unclear whether the circular process is able to cope with the speed. Basically, you already fail at the first point.

 

Example Scissor, Stone, Paper

EN_Baumdiagramm_11

 

2. Is there enough time to identify and interpret the stimuli?

Logically, this question must be negated. You don’t have any advance information. So, you cannot identify and interpret the stimuli at the necessary moment. Afterwards, you know what happened. The information is not lost. It was just not available. The process time is too long (remember the robot).

 

3. Is causality given?

This question can be answered by looking at the tree diagram. There are many possibilities due to the variable upcoming directional change. So, there is no causality. The missing stimuli (early patterns or traits) worsen the situation.

 

Tree diagram with variable speed and displacement

EN_Baumdiagramm_Scheibenwischer_02

 

4. Is it possible to differentiate between the alternatives?

This question targets any other available information that might help to reduce the number of possibilities. Are there any rules that forbid the usage of certain movements? Do the surroundings restrict movements, and therefore decrease the possibilities? In our case, the given rules of this simple game contribute significantly. It could be much worse if lifting the hand off the surface were allowed. Furthermore, the elbow is acting as a fixed rotation point that limits the possible pathway of the hand. The number of entries in the tree diagram would literally explode without these rules.

 

5. Is there a slack time for updating the internal model and the response selection?

There must be enough time between the initial information about what is going to happen and the target stimulus. Otherwise, the internal model cannot be updated and a response cannot be chosen. This time is difficult to determine, and it makes no sense measuring it directly in a training environment. It is easier to pay attention to whether there is short break between the initial information for the internal model and the target stimulus that triggers the selected response. Something is fishy if both happen at the same time. It indicates a choreographed scene.

This question cannot be answered for the sidewinder example. There is no advanced information and no causality. The slack time for this example is part of the lag between the directional change of your partner and your turn-around (see Anticipation, Part 3 for details). Using the transition area and its´ lag is one possible strategy to estimate the slack time.

 

Slack time between stimuli

EN_Reaktionskette_Antizipation_7

 

To be continued Anticipation, Windshield Wiper, Part 2

The post was published 18. July 2014 related to the category Miscellaneous and tagged with .