Skip to main content
lab-notes.aiPublic-service AI playbooks
Opportunity 14/17TransportTransport

Traffic Flow Management

Explore what a junction-level signal decision would need to know compared with the traffic counts and travel surveys currently published.

Inspect starter dataSet up the build partner
Published sources
2
Starter data
20 synthetic records
Domain build partner
build-traffic-flow
Last reviewed

From the draft strategy

Opportunity

The draft strategy names AI for traffic management as a potential public-service application, mentioning congestion prediction, signal optimisation, and spotting safety breaches, without naming a network or saying what the system would be trying to achieve.

Where the draft says it

Table 2 — potential public-service applications

Read the draft strategy source (opens in a new tab)

Source register

Research already done

Real published source

We reviewed these published sources so you do not have to start from zero. Each source shows what it covers, how it can be accessed, and why it matters.

2 sources

  1. Department for Infrastructure

    Open

    Traffic and travel information, incorporating the annual traffic census

    What it covers
    The yearly traffic report for Northern Ireland, combining the annual census — around 350 automatic counters and about 500 manual counts of traffic by vehicle type — with figures on how flows vary through the day and the year.
    Why it matters here
    It is the published count series this playbook's synthetic records imitate: vehicles by place and time band, with nothing recorded about any driver or any vehicle.

    Open the source (opens in a new tab)

  2. Department for Infrastructure

    Open

    Travel Survey for Northern Ireland

    What it covers
    A household survey of how and why people travel in Northern Ireland, reporting journeys by car, bus, walking, and cycling, and normally combining three years of responses to reach a large enough sample.
    Why it matters here
    It is where the bus and cycling shares in this example would have to come from, at a three-year regional grain rather than the junction and the hour a signal change actually happens at.

    Open the source (opens in a new tab)

Starter data

Starter dataset

Synthetic working data

Use 20 synthetic hourly counts across four junctions to explore signal timing without using a real traffic network.

AI-assisted research helped identify and interpret the published sources. We then created a small, non-sensitive synthetic dataset shaped by the information those sources expose. It is for exploration—not evidence, training, or operational decisions.

Records
20
Fields
6

Field preview

table view
  • id

    SYN-TF-01 · SYN-TF-02 · SYN-TF-03

  • junction

    Junction 01 · Junction 02 · Junction 03

  • hourBand

    07:00-08:00 · 08:00-09:00 · 12:00-13:00

  • vehicleCount

    1840 · 2310 · 1420

  • busShare

    4.2 · 3.8 · 5.1

  • cycleShare

    1.6 · 2.1 · 0.9

Inspect all 20 recordsView repository file (opens in a new tab)Download JSON (opens in a new tab)

How it was prepared

AI authored fictional vehicle counts and bus and cycle shares across five time bands for each junction.

Limitations

  • The junctions are numbers and the counts are invented, so nothing here describes traffic anywhere in Northern Ireland.
  • Four junctions are not a network: queues, incidents, and the effect of one junction on the next are exactly what a small table cannot hold.
  • Counting vehicles counts the people who are not in one badly, and the pedestrians who cross here do not appear in this file at all.

Clone and build

Domain build partner

Build with a domain-aware coding agent

Clone the repository, then ask your coding agent to load the skill that ships with it. The skill brings the opportunity, sources, starter data, known unknowns and constraints into the conversation before anything is proposed. It follows the open Agent Skills standard, so it works in Claude Code, Codex and any other agent that reads skills.

01 / Clone

git clone https://github.com/Hypership-Software/ats-us-nai.git

02 / Invoke

Claude Code
/build-traffic-flow
Codex
$build-traffic-flow

Conditions, not a checklist

Before you build

These are not footnotes. They are conditions that any responsible prototype would need to address.

  • Signal changes have wider effects

    A change to a signal plan reaches road safety, the people crossing on foot, emissions, and which streets end up carrying the traffic.

  • Counts under-represent what is not measured

    Counts under-represent whatever is least counted, so a route with no counter looks quiet and a mode with no survey looks absent.

  • Average delay can move the burden

    Average delay can fall while the burden moves onto pedestrians, buses, or one neighbourhood, and the headline figure will not show it.