FlowsBook a callSign inCreate free accountSign up
Workflows by industry

How warehouse operators plan staff shuttle buses for each shift

Employee shuttle bus route planning for a warehouse: choose pickup points, order them into loops, then check live road speeds at the hours the shifts really change.

Part of the guide Workflow automation

By the xMap team · · 7 min read

Employee shuttle bus route planning for a warehouse has three parts: decide where the bus stops, put those stops in the order that wastes the least road, and check the result against the traffic at the hour the bus actually runs. Most operators do the first two once and never do the third, which is why the 05:20 bus is early and the 13:20 bus is late on the same route.

Warehouses and distribution centres sit where land is cheap and public transport is thin, and their shifts start before the first train or end after the last bus. This article is for the operations and HR managers who run or buy staff transport for such sites. It covers pickup points, loops, the traffic check, a worked example and an AI employee that redoes the work as the roster changes.

Start with pickup points, not home addresses

A bus that calls at every front door is a taxi with forty seats. Group staff around a small number of shared pickup points instead: a station forecourt, a supermarket car park, a main junction with a safe place to wait. Three rules of thumb help.

  • A pickup point should be within a walk most people accept in the dark, and somewhere lit.
  • Each point should serve enough riders to justify the stop. One person at a far-off point is a conversation, not a route.
  • Keep a point per shift only if that shift uses it. The night shift's map is rarely the day shift's.

There is a privacy reason too. A table of pickup points with a count of riders at each is all a routing tool needs. Home addresses can stay in the HR system.

Order the stops into loops

With the points fixed, the question is the order to visit them and how many buses to use. Two limits shape the answer: seats on the bus, and the longest ride you will ask of the first person to board. A loop that fills the bus but keeps that person on board for 80 minutes will lose them to another employer.

The Route planner block puts the places that reach it in the best order to visit from a start you pick, along the roads. It handles up to 50 stops in a route, can share the stops between a number of vehicles, from two to ten, and gives each stop its place in the order, the distance and time from the stop before, and the route's totals, with a directions link the driver can open on a phone.

Two details matter for a staff bus. First, every route begins at the start you pick, and with Come back to the start switched on it ends there too. So plan the morning run as a loop from the warehouse gate: out empty, round the pickup points, back full. The evening run is the same loop with riders getting off. The planner does not fix a different final stop, so a one-way run from a bus garage to the warehouse is better planned as this loop. Second, its times are travel times for usual traffic, and the minutes spent at each stop while people board are not added. Add them yourself, a minute or two a stop.

Check the roads at the hour the shift changes

A route planned on usual traffic is a draft. The test is what the roads do at 05:20, 13:20 and 21:20, or whenever your buses are moving. Traffic Flow measures the roads around each place at the moment a flow runs: the current speed, the speed when the road is free, a jam factor from 0 to 10, and the congestion as a percentage. Because it is live, a flow scheduled just before each shift change tells you how that hour looks, and a week of such runs written to a sheet tells you how it usually looks.

Be clear about what that is. It is speed, not volume. It does not count cars and has no built-in peak hours, so you build the picture by running it at the times you care about. Hourly car counts going back years exist only for roads in the United Arab Emirates, as a regional dataset. Elsewhere, your own week of readings is the evidence.

Run it on the pickup points and the warehouse gate. The stops whose speed falls furthest below free flow are where the timetable needs slack, and sometimes where the order should change: visiting a congested junction first, before the school traffic, can save more time than any shorter path.

A made-up example: three shifts, two buses

Picture an invented distribution centre outside Guadalajara with 140 staff on shuttles across three shifts and 14 pickup points. Every figure here is made up to show the method.

LoopStopsPlanned timeAt 05:20At 13:20
North bus846 min44 min63 min
South bus638 min37 min41 min

Invented loop times. Planned time is for usual traffic before boarding minutes are added.

The early run is fine. The afternoon run on the north loop is 17 minutes over, and the readings show why: two stops near a market street move at a third of their free speed after one o'clock. The manager makes two changes. The afternoon north bus leaves 15 minutes earlier, and its order is reversed so the market stops come first. A third change waits for a second week of readings: moving one pickup point 300 metres to a wider road.

Mistakes that cost riders

  • One timetable for all shifts. The same loop takes different times at different hours. Publish a timetable per shift.
  • Planning on the roster of last spring. People join, leave and move. A stop with two riders left is a candidate to merge.
  • Forgetting boarding time. Eight stops at two minutes each is a quarter of an hour that no routing figure includes.
  • No plan for a closed road. A closure on the only bridge makes the whole loop late. Check for incidents before the bus leaves, not after people are waiting.
  • Measuring once. One morning's reading is an anecdote. Keep at least two weeks before changing a route.

An AI employee that keeps the plan current

This is recurring work with a clear timetable, which is what the field operations guide describes an AI employee doing. In xMap Flows it is two small flows.

The weekly planner runs on Sunday evening.

  1. A Schedule, weekly.
  2. My data reads the pickup-point table, kept in a sheet HR updates: name, address, shift, riders.
  3. A Filter per shift, dropping points with no riders this week.
  4. Route planner, starting at the warehouse gate, shared between your buses, coming back to the start.
  5. An Email to the transport supervisor with each bus's stops in order and its link, and a row per stop in a sheet.

The shift check runs 40 minutes before each shift change, using the Schedule's several-times-a-day setting.

  1. My data reads the same points.
  2. Traffic Flow measures the roads around each, and Traffic incidents looks for closures and roadworks near them.
  3. A Branch on the jam factor: points above your line go one way, the rest are only logged.
  4. A message on WhatsApp or Slack to the supervisor naming the slow stops and any closure, so the bus can leave early or skip to the next point.

What lands in the inbox: on Sunday, next week's loops; before each shift, either nothing or a two-line warning. For operators selling shuttles to employers, the template Business parks far from stations finds sites with the same problem.

Questions

How many pickup points should one staff bus serve?

Few enough that the first person to board is not on the bus for much longer than they would accept by car. For most sites that means a handful of stops per bus, not a dozen.

Can the route end at the warehouse and start somewhere else?

The Route planner starts every route at the place you pick and can return there. Plan the run as a loop from the warehouse gate; the same loop serves the journey in and the journey home.

Does the traffic data show rush hours for my roads?

Not as a ready chart. Traffic Flow measures speed at the moment it runs, so schedule it at your shift times and keep the readings. Hourly car counts over past years exist only for roads in the United Arab Emirates.

How often should the routes be replanned?

Whenever the roster changes enough to empty or fill a stop. A weekly run from a table that HR keeps current is a sensible default.

Do I need staff home addresses to plan the routes?

No. A table of shared pickup points with the number of riders at each is enough, and it keeps personal addresses out of the planning tool.

Templates that do this

Read next

More on workflows by industry · 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.