The examination of Los Angeles in previous articles brings to the surface issues regarding movement of various entities in various scales. So far, we’ve examined the movement of dense sprawl, movement from fringe to core, movement through freeways and water infrastructure, and programming movement. How do we control and choreograph movement of such large density?
To understand efficient movement of densities, we can look to swarm movement patterns and swarm intelligence found in biology for guidance. Swarm intelligence can be thought of as a system of a dense and populous agents that interact locally with other agents and their immediate environment. There is no centralized system of movement that controls the agents as a whole. Instead, these local and individual decisions open up the possibility for self-organization of the whole. Coherent, global behavior emerges from individual, local decisions.
The Boids algorithm simulates swarm behavior by hypothesizing that agents follow the three rules of cohesion, alignment, and separation when interacting locally. In cohesion, agents move towards the average position of its neighbors. In alignment, the agents aligns its steering direction towards the average direction of its neighbors. In separation, the agent moves away from the neighbors if they are too close. These interactions occur between different agents depending on their relative distance from each other.
The movement of bacteria is particularly interesting because they communicate with other bacterial agents with physical contact, they are adaptable in their decision-making, and they become uniquely and aggressively coordinated when faced with prey.
Bacterial agents have pili, hairlike appendages on their surface, that attach to other agents as a way to locate them. Pili also allows for agents to maintain a certain distance from each other. This physical sensory feature allows for decision-making to be more intuitive and less thought-based.
Bacterial agents also efficiently adapt their decision-making when they receive positive or negative feedback. When individual agents find a more efficient path of movement, they tend to tune out the other agents around them, acting more autonomously, with neighboring agents following suit. On the other hand, when individual agents are failing, they tend to increase their interaction with their neighbors and adapt their movement strategy to that of their neighbors.
Once faced with prey, bacterial agents act with even more purpose. The swarm moves in parallel ripples towards the prey like waves hitting the shores of a beach. Agents line up perpendicular to the axis of the ripple and move back and forth along each wave, targeting the prey along a swarm-manufactured, moving grid.
Swarm intelligence teaches us that movement can become automatic based on simple rules and extensions of these rules. Behavioral modeling tools can help us simulate the movement of swarms and add complexity to their movement logic. We can analyze their occupation of designed environments. We can track their patterns of circulation, occupation, and interaction with space. Using this technique, the functionality of crowds and the space they occupy (or don’t occupy) is analyzed intelligently through the lens of movement.
Can we creatively automate the movement of people in Los Angeles? Or the movement of cars? Or the movement of populations?