How do you work out how much work fits into a field crew's shift?

Count in minutes, not in job cards. From the length of the shift subtract travel time between sites, preparation and wrap-up, breaks and slack for deviations. Divide the remainder by the standard duration for the job type in question, taken from actual recorded start and finish times rather than from impressions. If the spread within a type is too wide, split it into subtypes.

Generalist crews or specialist ones?

Generalists are easier to plan: any job fits any team, so workload evens out on its own. Specialists close complex types of work faster and to a higher standard, but their idle time is not offset by someone else's flow of work. In practice companies keep a core of generalist crews and a few specialist ones for work that needs particular certifications and equipment.

How much of the shift should be kept for urgent calls?

There is no universal share — it comes out of your own statistics. Look at how many urgent jobs arrived by day of the week over several months, and at what hours. The reserve is set aside either as an on-call crew or as free time in everyone's shift. A plan loaded to a hundred per cent breaks at the very first deviation.

How is crew planning different from job dispatching?

Crew planning is the upper loop, run weekly and monthly: how many crews go out, in what composition, across which areas and with what reserve. Dispatching is the operational decision of the day, made inside the capacity already created. If reschedules happen every day rather than occasionally, the problem is almost always in the upper loop, and shuffling tasks in the evening will not solve it.

Where do you start if the roster lives in a spreadsheet?

With the data: normalise your addresses, gather real duration standards, build a skills and certifications matrix, write down the priority rules and the cut-off for urgent jobs. Then record your current metrics and run a pilot with one team, comparing the result with the baseline a month later.

Field crew scheduling: shift capacity and rosters

Field crew scheduling is more than laying tomorrow's addresses out across people. First you decide how many crews go out on a shift and in what composition, how much work they can physically get through, and how much slack is left for emergency calls. A mistake at that level can't be cured by clever dispatching: if the plan holds more work than fits into a day, evening reschedules are inevitable. Let's look at how to calculate shift capacity, put crews together, run a weekly roster and tie it to the geography of your sites.

Crew planning and job dispatching are different problems

In field service it is easy to blur two processes that live on different horizons. Dispatching is an operational decision: who takes this particular job today. Crew planning is a decision for the week and the month ahead: how many people go out, who works with whom, which areas and job types they cover, and where slack is kept for urgent calls.

The difference shows in the symptoms. If failures are isolated and tied to specific addresses, the problem is in the daily assignment. If every day ends with jobs pushed to tomorrow and the peak always lands on the same weekdays, it is the upper level that is broken: shift capacity does not match the flow of work. No amount of shuffling tasks between people in the evening makes up that shortfall.

In practice this means two separate control loops. The upper loop sizes capacity: how many crews, what they consist of, the pattern of shifts, the split across service areas. It is reviewed once a week and rests on demand statistics — how many jobs of each type came in by day, and how many of them were urgent. The lower loop works inside the capacity you have already committed to, and solves the problem of the day.

When the two are not separated, the service manager spends every morning patching holes that should have been closed last week by a rostering decision. Formally it looks like a dispatcher problem; in substance it is the absence of capacity planning.

Shift capacity: how much work actually fits into a day

Capacity answers one question: how many jobs a crew closes in a shift without heroics. Count it in minutes rather than in job cards, otherwise a survey and a full installation end up weighing the same in your plan.

From the length of the shift you subtract, one after another:

Travel time between sites. In dense urban areas driving eats a share of the shift comparable to the work itself, and it is exactly what spreadsheets tend to leave out.
Preparation and wrap-up. Picking up materials from the warehouse, loading the van, paperwork and the photo report after the visit.
Lunch and statutory breaks. Always subtract them, otherwise the plan is built on your staff's unpaid time.
Slack for deviations. A part of the shift deliberately left empty for emergency calls, jobs that overrun and return visits.

What remains is the working time available for planned jobs. Next you need duration standards by type: how long a survey, an installation, a warranty repair or scheduled maintenance takes on average. Take the standard from actual data — from completed jobs with recorded start and finish times, not from an engineer's impression. It pays to look at the spread as well as the mean: if a routine job takes anywhere from forty minutes to three hours, planning by the average is pointless and the type needs splitting into subtypes.

Capacity at bottlenecks is counted separately. If work on a particular piece of equipment can only be done by two crews out of eight, those two become the constraint for the day — the total free capacity of the service has nothing to do with it.

Crew composition and the skills matrix

Composition determines which jobs a crew can take at all. So before you build the roster, it helps to draw up a matrix: employees down the side, job types, certifications and equipment across the top, and in each cell — can do it alone, can do it in a pair, cannot do it.

The matrix exposes weak points immediately. If a single person is ticked for some type of work, their holiday, sick leave or resignation halts that line of business entirely. Close those points in advance: training, sending a newcomer out alongside an experienced specialist, spreading certifications wider.

When you put crews together you usually have to choose between two strategies. Generalist crews take any job: they are easier to plan, because any task fits any team, but training people costs more and takes longer. Specialist crews close their narrow type of work faster and to a higher standard, yet planning gets harder: each has its own stream of work, and one sitting idle is not offset by another being busy.

The workable compromise is a core of generalist crews plus a few specialist ones for complex work. The main flow is then distributed freely, while rare types go to the people who genuinely do them.

It is also worth fixing the pairings. A team that has worked together installs noticeably more predictably than one thrown together in the morning, so rotating people at random every day to even out workload usually costs more than it appears to.

The weekly roster: planned work and room for urgent jobs

A roster for field technicians is built a week at a time and rests on the shape of demand. The first thing to see is how jobs spread across the days: there are almost always overloaded days and empty ones, and the first improvement comes not from hiring but from moving planned work into the quiet days by agreement with the customer.

The frame of the weekly roster is assembled step by step:

The availability calendar. Shifts, days off, holidays, training, scheduled vehicle servicing — everything that makes someone unavailable.
Demand by day and job type. The expected volume in minutes rather than in job cards, allowing for seasonality and commitments already made.
Capacity to match that demand. How many crews and of what composition you need out on each day, and where scarce skills fall short.
A reserve for urgent calls. Either an on-call crew or a share of every shift kept free.
A cut-off rule. Up to what hour an urgent job goes into the current day, and after which it moves to the next one with the customer notified.

Planned jobs are then laid out across the days with geography in mind rather than wherever they happen to land: work in the same district is worth grouping into one day, which keeps the route within that day short. This is the step where weekly planning meets routing.

Agree separately on what counts as a change to the roster and who approves it. If any manager can drop a job into a full day, planning turns into decoration.

Assigning technicians to addresses

Once composition and roster are settled, the daily part begins: assigning technicians to addresses and building the order of visits. Three groups of conditions are at work here.

Hard constraints rule out impossible options: the required qualification and clearance, the time window agreed with the customer, the crew's working hours, access to the site, vehicle capacity for materials. Priorities resolve conflicts: an emergency displaces planned work, a contractual deadline beats convenient geography, a return visit stays with the same team. Optimisation criteria pick the best of the acceptable options: total mileage, the number of jobs closed, evenness of workload in minutes.

By hand this only works at small volumes. As soon as you have dozens of addresses, several crews and time windows, the number of possible visiting orders explodes and a person stops seeing even a decent solution, let alone the best one. The classic sign that you are at the limit: laying out the morning takes more than an hour, and by lunchtime it has to be rebuilt from scratch.

A practical yardstick: how long it takes to move three jobs from one crew to another when a van breaks down. If that is half an hour of phone calls, your allocation rests on the telephone, and on a peak day it will predictably fall apart.

Moving from a spreadsheet to a system

Make the move gradually. Data first: addresses with coordinates, real duration standards by job type, the skills and certifications matrix, the availability calendar. Then record where you stand today — jobs per crew per day, share closed on the first visit, share completed within the agreed window, mileage per job. Without those baseline figures you will not know what changed after automation. Next, a pilot on part of the flow: one team, one city, a couple of weeks running alongside the usual process.

What changes as a result. Jobs are accepted and allocated to field staff in a single screen, routes are visible on a live map, and the visiting order is calculated by an OR-Tools optimisation engine that respects delivery windows and constraints. The technician works in a mobile app with photo reports and checklists — a mobile web interface is available too, with no app to install — completion is confirmed on the spot, and the customer sees statuses in their own portal. How this is set up for field teams is described on the Field service management page.

You do not have to retype the data either: itlogist exchanges with 1C, AmoCRM, Bitrix24 and Excel, so jobs reach the plan from the sources you already use. Rollout takes 7 days, with no long-term contracts; the product is built for teams of 5 to 100 field staff. You can test a new planning scheme on live work and compare it with your recorded baseline a month later.

Field crew scheduling: the roster, crew composition and shift capacity