DALI TCS Models & Methods

System Model

(This anchors everything.)

What to include:

  • Network of signalized intersections
  • Each intersection modeled as an autonomous agent
  • Local sensing inputs (queues, arrivals, phases)
  • Neighbor relationships among intersections

Example framing (not final text):

We model the urban traffic network as a graph of signalized intersections, where each node corresponds to an autonomous control agent responsible for local signal timing decisions.

Agent Model

(This is where MAS reviewers look first.)

What to specify:

  • Agent state (local traffic conditions, timing constraints)
  • Perception model (what information is available locally)
  • Action space (phase selection, timing adjustments)
  • Local objectives (delay reduction, fairness, stability)

This distinguishes DALI from:

  • centralized optimizers
  • purely reactive actuated control

Communication & Coordination Model

(This is one of DALI’s defining contributions.)

What to include:

  • Direct agent-to-agent communication
  • Neighborhood scope (which agents communicate)
  • Information exchanged (summaries, intent, constraints)
  • Timing and frequency of communication

You should explicitly emphasize:

  • Real-time
  • Peer-to-peer
  • No central coordinator

This is where your “direct agent-to-agent communication” claim belongs.

Decision-Making & Control Strategy

(Methods, not code.)

What to include:

  • Distributed decision process
  • Adaptive strategy under dynamic demand
  • Cooperative mechanisms (not competitive or isolated)
  • How local and shared information influence decisions

This can stay high-level, with references to papers for details.

Assumptions & Constraints

(Signals rigor and honesty.)

What to include:

  • Sensing assumptions (baseline vs richer sensing)
  • Communication reliability assumptions
  • Timing constraints imposed by signal controllers
  • Infrastructure compatibility

This builds trust with technical readers.

Design Rationale (Optional but Powerful)

(Why these choices?)

What to include:

  • Why decentralization over centralization
  • Why agent-to-agent coordination
  • Why incremental deployability mattered
  • Tradeoffs considered

This is where SCALA’s engineering judgment shows.