Application Ideas for Self-organizing Systems

Autonomous Transportation Systems

Pipe systems for transportation of goods
Autonomous Robots transport goods from A to B
Goal:
- Agents have to find their way to these pipe systems themselves
- Agents have to move inside the pipe system from A to B
- Problem: if many agents in the system how to organise them to avoid jams, …

From top-down system to bottom-up system
Decentralised
Constraints, zero-tolerance against error
Optimise the route
Agent in the plane knows about other neighbouring planes and applies repulsion
- Identifies the route (on the air)
- Avoid other planes (on the air) – take regulations into account
- Improve quality of service (time departure, ask for priority to leave the airport)
Reasons
- Many points of failures
- Humans involved – air traffic controller

Put software agents into stock markets
Goal:
- To make market more stable
- Avoid crashes
- To be more in line with economic theory (rational behaviour, faster, act in zero time)
How (Alternatives)
- All the trading is done by agents
- Trade based on trust
- Build a network of trust, help identify grey-zone
- Population of agents that crawls the market and looks at it in other ways and identifies whether or not product/entity is trustful
  - Agents rely on human-like trust (identify healthy from unhealthy products)
  - Agents rely in facts on not on emotion or imitation
  - Agents identify risks
Open questions
- How to mix them with human people
Acceptance?

Use embedded or integrated sensors to sense mass movement
Use mobile self-organizing nodes (e.g., drones) to get more information in areas of interest
Reroute persons optimizing time to exit, time to find a free space in a train, etc.

Using cooperating smart homes equipped with cameras, facilities to autonomously switch on light/TV
Apply pattern recognition to identify theft attempts (e.g., persons spying around)
Self-organization mechanisms are required to adapt to changing patterns and situations