After going through numerous papers and articles on Human-Swarm Interaction (HSI), I want to present my idea on where the future research of HSI should be headed.
The field of Human-Swarm interaction is trying to mimc the co-ordinated task done by biological animals (bee, ants and others) to robots. We researchers see a great bright human future by developing intelligent and autonomous bio inspired collections of robots working for a common goal. We think that bees and ants behaviors are the best for research on human-swarm interaction.
But why not consider human ? We humans are able to form complex structures and solve abstruse problems by co-ordinating locally as well as globally. What makes the study interesting where humans behaviour are modelled for swarm ? We humans are the only creature who believes on ideas and fiction. The daily tasks we humans perform is mostly based on belief and fiction. i.e. we do tasks that doesn’t benefit us directly but we believe our actions will somehow be helpful in near future. If we can model this concept of ideas, belief and fictions and build robots based on this model, then we can minimize lots of research problems like fan-out, optimal balance between span and persistance.
Moreover, we can completely change the way we interact with swarm using the idea described above. We no longer need to depend on the attractor to govern or control the swarms behaviour. We only need to control the central idea or the virtual environment on which the swarm is working.
Lets consider an example: Money seems omnipresent to humans but is fictious and our whole world is driven by money. Now lets try to bring this fictitious idea of money in case of swarm interaction. Lets try to model the idea of money by some function. As money for humans is never ending and can always be aquired more with more profits and hardwork. Money in swarm world can be model as a function trying to find largest prime number. Similar to money, there should have exist largest prime number than current one and is always computable to given computational resources.
Now, the swarms will exist in a environment where the ficitous goal is to find the largest prime number. But the real goal could be disaster recover, survellence or any other task. For controlling the swarm, we just can change the functional parameters of the environment and don’t need to communicate with the attractor. This method doesn’t require to find the optimal balance between span and persistence. Also, this method will allow humans to control many human inspired swarms.
The method that I envision might take use to a different world of human-swarm interaction. Below are the excellent papers that inspired me to think of this new approach.