The morning puzzle
Every morning, or late every afternoon, someone sits with a list of jobs, a map and a spreadsheet. They work out who goes where, in what order, fitting in the customer who can only do after two, the job that needs a gas-qualified engineer, and the engineer who has to finish early on Thursdays.
It takes a good chunk of time. It depends on one person's local knowledge. When an urgent job comes in at ten, the plan is redone on the phone. And even on a good day, engineers are crossing each other's paths across town.
Why it is still done by hand
Route planning looks like a map problem. It is really a constraints problem: time windows, skills, van stock, working hours, breaks, priorities and job lengths that vary. Simple mapping tools reorder stops for one driver but ignore most of that, so planners give up on them.
Full field service packages include scheduling, but they often expect you to move all your jobs, customers and processes into them first. For a business with a working job system, that is a lot to change just to get better routes.
And the planner's knowledge matters. They know which customer is always late to answer the door and which engineer is best with a certain client. Any tool that ignores that gets overruled, and rightly.
What hand planning costs
| Cost | How it shows up |
|---|---|
| Planner time | A skilled person tied up each day doing arithmetic |
| Extra driving | Fuel, vehicle wear and time on the road |
| Fewer jobs per day | Capacity lost to travel rather than work |
| Missed time slots | Customers waiting or rebooking |
| Single point of failure | When the planner is off, the plan suffers |
| Slow reaction to changes | Urgent jobs slotted in wherever there is a gap |
Extra driving is the one most businesses underestimate, because it is spread across every van every day and never appears as a single number anyone looks at.
How we automate it
- Data in. We take each day's jobs from your job system through its API: address, customer time window, expected duration, required skills and priority.
- Engineer details. Start and finish locations, working hours, breaks, skills and qualifications, and van stock where relevant.
- Optimisation. We send this to a route optimisation service, such as Google's route optimisation tools or an open-source solver, which proposes routes that respect the constraints and reduce travel time.
- Planner review. The planner sees the proposal on a map and a timeline, drags jobs between engineers where their knowledge says otherwise, and locks anything that must not move.
- Publish. Approved routes go to engineers' phones with jobs in order and navigation links.
- Re-plan on the day. When an urgent job or a cancellation arrives, the planner asks for a new suggestion for the remaining day, with completed jobs and locked jobs kept in place.
- Customers can be sent an arrival window once routes are published
- Actual versus planned times feed back to improve duration estimates
- Constraints are stored as settings the planner can edit, not buried in code
What mornings look like
The planner starts from a sensible draft rather than a blank map. Their time goes on the judgement calls, not the arithmetic. Engineers get a route that makes sense, and customers with a time window are more likely to see the engineer within it.
When the planner is off, someone else can produce a reasonable plan, because the rules are in the system rather than in one person's head.
Urgent jobs become less disruptive as well. Rather than ringing round to find who is nearest, the planner asks for a revised plan for the rest of the day and sees which engineer can take the job with the least knock-on effect on everyone else's customers.
Is your planning like this?
- Routes are planned by hand with a map or spreadsheet
- One person holds the knowledge of how to plan well
- Engineers cross paths or double back during the day
- Urgent jobs throw the whole plan into disarray
- Skills, time windows and working hours are juggled mentally