What is it about?
In simulations, humans are represented with agents that know the shortest path to their goal. Humans don't always take the very best path, just a good one. How come? One neglected aspect is that humans don't remember the environment perfectly. We use findings from cognitive science to model the bias that humans have in remembering spatial relations and give it to agents. Our agents move more like humans and stop using the perfect path but still solve tasks reasonably well.
Featured Image
Photo by Aedrian on Unsplash
Why is it important?
Agent-Based Modelling is becoming the bread and butter for research on cities and human spatial behaviour. Sticking to agents that work like automata produces results that do not reflect reality. Getting agents to have a more human cognition will allow us to more effectively predict human behaviour on large scales and make these simulations more useful.
Perspectives
Since the A* algorithm was developed, agents nearly stuck on the same way of solving way-finding and routing. Our contribution updates this approach and gives new impetus to the question of how to reasonably design agents to conduct research with real-world impact.
Jascha Grübel
ETH Zürich
Read the Original
This page is a summary of: A cognitive model for routing in agent-based modelling, January 2019, American Institute of Physics,
DOI: 10.1063/1.5114245.
You can read the full text:
Contributors
The following have contributed to this page
