75

Technical Research |Simulations Ideology

74

Technical Research |Simulations Ideology

This concept is especially important when simulating humans due to their inherent complexity. The human brain is a complex object and to this day is still not fully understood, therefore, simulating human behaviors perfectly is beyond the current capabilities of collective human knowledge. As stated by Delaney and Vaccari in Dynamic Models and Discrete Event Simulation: “one approach with coping with such difficulties is to relax the requirements on the model so that approximation can be employed even though they yield fewer and/or less accurate results”[8] This means that simulations can vary in levels of complexity, ranging from simple real time calculations for approximating visualizations to high accuracy scientific models. Banks et al. also acknowledges this, stating:

“The art of modeling is enhanced by an ability to abstract the essential features of a problem, to select and modify basic assumptions that characterize the system, and then to enrich and elaborate the model until a useful approximation results. Thus, it is best to start with a simple model and build towards greater complexity. However, the model complexity need not exceed that required to accomplish the purposes for which the model is intended. Violation of this principle will only add to model-building expenses. It is not necessary to have a one-to-one mapping between the model and the real system. Only the essence of the real system is needed.”[9]

It should be noted how this aligns with Jason Gregory’s description of video games back in Chapter 1.3, which is a good indication of the validity of this approach when the time comes to integrate this model within a Game Engine. With this in mind, it can then be determined that a crowd simulation for architectural visualization purposes—that can also be evolved by architects—would benefit more from simplicity and speed, rather than perfect accuracy from the beginning. This is already an improvement over not simulating at all, and since the behavior of people is not absolute and exact, an approximation is likely sufficient to gauge whether or not a space is working properly in a crowd situation. It is not required to figure out all the nuances of the system, but instead focus on determining simple agents that can utilize emergent behaviors from a bottom-up approach. These agents would then act upon simple embedded rules to move around the space, and if need be, higher accuracy can be generated by adding additional simple rules to these agents. One does not need to simulate behaviors perfectly, but rather account for enough variance that the system as a whole is realistic enough to provide information and believability to inform the design process. In doing so, it makes the creation process much simpler; instead of creating artificial intelligence, this simulation can be based on simple rules to give the illusion of distributed intelligence. (Fig. 2.1.7 - 8)

Rule 30 as introduced by Stephen Wolfram, 1983

What is compelling about the phenomenon of emergence in the context of crowd simulations is how one can utilize relatively simple logic to generate complex behaviors.

From Eric W. Weisstein, “Rule 30,” Wolfram Math World, accessed December 25, 2019, http://mathworld.wolfram.com/Rule30.html.