Using live traffic to plan deliveries and service visits

Use live traffic in two places: at the depot, to decide when to send vans out, and around the day’s stops, to decide their order and the arrival windows you promise. Traffic Flow measures the roads around each place at the moment the flow runs; Route planner orders the stops; a Branch on the jam factor turns the reading into a hold or a go.

This article is for dispatchers, field service managers and small delivery operators. It explains why a route plan alone is not enough on bad mornings, sets out three flows of rising ambition, and warns about the habits that make live data less useful than it should be.

Why a route plan alone falls short

Route planner works out the best order of stops and the minutes between them for usual traffic. That is right most days. On the morning a motorway junction locks up or a stadium empties, the usual minutes are wrong, and a technician promised for 10:00 arrives at 11:15.

Live speeds cannot reorder the world, but they tell you early. A dispatcher who knows at 06:45 that the roads outside the depot are at a jam factor of 8 can hold departures for half an hour, call the first customers, or swap the first two stops.

Flow one: hold or release at the depot

1. Start with a Schedule set to every 30 minutes, with active hours from 06:00 to 10:00 on weekdays.
2. Add My data with your depots.
3. Add Traffic Flow with Roads set to Within 1 km, to catch the main exits.
4. Add a Branch: jam factor 4 or more (Moderate and worse) goes one way, the rest the other.
5. On the Yes side, a Slack message to the fleet channel: hold departures from this depot, with busiest_road and the speed against free flow.
6. Leave the No side empty, or post a short "roads clear" note there if your dispatchers want one.

The ready-made version is the template Road speeds outside my depots, every hour, which lists the most jammed depots first in a short note for dispatch.

Flow two: today’s stops, checked before the round

For service visits, the stops change every day. Feed the day’s jobs from a table or a New rows trigger on the booking sheet, run Traffic Flow on the roads next to each stop, and pass them to Route planner with your start and the number of technicians. Each stop arrives with its visit order, the minutes from the stop before, and its jam factor beside it.

A made-up example from Johannesburg: an air-conditioning firm has three technicians and 24 jobs. The 07:00 run shows four jobs in one business district with a heavy jam factor. The dispatcher moves them to the afternoon and gives those customers a later window before anyone sets off. In Madrid, a made-up pharmacy wholesaler does the same for its first drops around the centre and finds the inner ring free flowing but one bridge heavy; only one van changes its order.

Flow three: your own record of the worst hours

Live traffic data has no peak hours or direction split of its own: it measures now. If you want to know when the roads around your depots are worst, record it yourself. Run Traffic Flow every 30 minutes for a few weeks and write a row each time into Google Sheets with the depot, the time, the speed, the free-flow speed and the jam factor. After a month you have your own picture of every depot’s bad half hours, weekday by weekday, built from real measurements.

That record is often more useful than any single morning’s reading: it tells you to move the standard departure from 07:00 to 06:30 at one depot and to leave it alone at another.

Habits that waste live data

• Checking at the wrong time. A reading at 05:00 says nothing about 08:00. Schedule the flow for the hour the decision is made.
• Alerting on every change. Use a threshold (a jam factor of 4, or 7 for heavy only) and send one message when it is crossed.
• Reading speed without free flow. 35 km/h is fine on a city street and a jam on an expressway; the jam factor and congestion already account for it.
• Expecting car counts. Traffic Flow measures speed, not how many vehicles pass. Car counts by hour exist only for roads in the United Arab Emirates, in Regional data sources.
• Forgetting closures. A closed road shows in roads_closed. For closures and roadworks across a whole area, add Traffic incidents (see road closure alerts for field teams).