FlowsBook a callSign inCreate free accountSign up
Roads and traffic

Where to put a delivery hub that serves two cities

Halfway on the map is rarely halfway by road. How to compare candidate hub sites between two cities by travel time, live road speeds, incidents and who can be reached.

Part of the guide Location intelligence automation

By the xMap team · · 5 min read

The best place for a delivery hub between two cities is the one that reaches the most of your deliveries in both, within your promised time, at the hours your vans actually leave, not the point halfway on a straight line. Shortlist sites near the main road linking the cities, then compare them on how far a van gets from each in a fixed time, how fast the roads around each move at your dispatch hours, how often those roads close, and how many people or customers sit within reach on each side.

This is for logistics and operations teams at couriers, wholesalers, e-commerce brands and service companies weighing one shared hub against two separate ones. It walks through the questions, what live road data does and does not show, a worked example, and an AI employee that keeps measuring the shortlist until the lease is signed.

Why the midpoint is usually wrong

Draw a line between two city centres and its middle often lands in fields, on the wrong side of a river, or beside a road with no junction. Even when it lands on the motorway, three things pull the real answer away from it.

  • The cities are not the same size. A hub that serves a city of several million and a town a tenth of its size should sit closer to the big one, where most drops are.
  • Roads are not symmetric. One city may have a ring road that lets vans in quickly; the other may have one congested bridge.
  • Time of day matters. A site that is twenty minutes from both centres at 11:00 may be fifty from one of them at 08:00, when the first wave leaves.

Four questions for every candidate site

  1. How far does a van get in a fixed time? Draw the area reachable by car from each site. Compare how much of each city falls inside it.
  2. How do the roads move at dispatch time? Measure the current speed against the free-flow speed, and the jam factor, on the roads around each site, at the hours vans leave and return.
  3. How reliable are those roads? Count closures and roadworks near each site and along the link between the cities over several weeks.
  4. Who is within reach? Count your customers, or residents where population data exists, inside each site's reachable area, split by city.

Reading the road figures

Free-flow speed is how fast traffic moves when the road is clear; current speed is how fast it moves now. The gap between them, shown as congestion in percent and the jam factor from 0 (free) to 10 (standstill), is what costs your drivers time.

What you see at 08:00 over four weeksWhat it suggests
Low jam factor every morningVans leave on time; a reliable site
High jam factor on Mondays onlyPlan a later first wave on Mondays
Moderate most days, standstill some daysLook at incidents: a recurring closure or roadworks
Free roads near the site but heavy at the city edgeThe site is fine; the problem is entering the city, which every site shares

How to read repeated morning measurements around a candidate site.

A worked example

Here is a made-up case, using the one place where car counts exist. A parcel company wants one hub to serve both Dubai and Abu Dhabi and has three sites on its shortlist along the road between them.

For each site it draws the area a van reaches in 15 minutes and checks which side of the corridor that area opens onto. It runs a live road reading at 06:30, 08:00 and 17:30 every weekday for a month, and keeps every road closure found within 5 kilometres. Because the sites are in the United Arab Emirates, it also reads the cars counted hour by hour on the main road near each site over the last 12 months, which shows that the site nearest Dubai sits beside a road whose busiest hour falls exactly when the morning vans would leave.

The middle site has the calmest mornings and the fewest closures, while the site nearest Dubai has the most customers within reach. The team chooses the middle site and plans a small parcel point inside Dubai for the earliest deliveries. For a pair of cities elsewhere, such as Lyon and Saint-Etienne or Toronto and Hamilton, the method is the same without the car counts: travel areas, live speeds measured over weeks, incidents and the customers within reach.

An AI employee that measures the shortlist

A site decision takes weeks. An AI employee can collect the road evidence on schedule while the team negotiates.

  1. Schedule: several times a day, at your dispatch and return hours, on weekdays.
  2. My data: the candidate sites with their positions.
  3. Traffic Flow: the roads within 500 m of each site: current speed, free-flow speed, jam factor, congestion and the busiest road.
  4. Traffic incidents: closures and roadworks around each site.
  5. Google Sheets: each reading added as a row, so the weeks build up.
  6. Email: a short note when a site's roads are heavy or closed.

A second, monthly flow handles reach: Travel area by car from each site, then Match with my table against your customer list or Demographics where population data exists, so each site gets the customers or residents inside its area. In the United Arab Emirates, a Regional data step with Road traffic history adds the cars counted on the main road near each site.

What lands in the inbox is a growing record per site: how its roads behaved at each dispatch hour, every closure, and who it reaches in each city. The decision meeting starts from evidence instead of a map and a ruler.

Questions

Can I see how many cars pass each candidate site?

Only in the United Arab Emirates, through the Road traffic history dataset in Regional data, which counts cars hour by hour since January 2021. Elsewhere, Traffic Flow gives live speeds and the jam factor, not car counts.

How far can a travel area reach?

Travel area draws the area reachable within a travel time or a distance by road, by car, on foot or by bike. Today the longest it draws is 15 minutes or 10 km, so compare sites on what each reaches in that time.

Is one hub better than two for two cities?

It depends on how many drops each city has and how reliable the link road is. If one site reaches most customers in both within your promised time at dispatch hours, one hub works; if not, a small second point helps.

How long should I measure the roads?

At least four weeks at your real dispatch hours, so you see normal days, bad days and the pattern by weekday.

Templates that do this

Read next

More on roads and traffic · Every article · Templates

Put it to work

An AI employee in xMap Flows does this on live data, on its own timetable. Start from a template or describe it in plain words.