Proseware: Field Scheduling That Cannot Fill a Slot
Proseware schedules 60 technicians for about 400 service appointments a month using Dynamics 365 scheduling. Most jobs need one technician, but about a fifth need two technicians with different skills plus access to a customer lift.
Dispatchers using Schedule Assistant frequently get no results for the two-technician jobs, then book manually and create double bookings. Travel time between sites is not reflected, so technicians arrive late and later appointments slip. Some technicians are booked outside their territory and refuse the job. Proseware wants the dispatchers to trust the scheduling tools rather than working around them.
1. How would you model the two-technician, two-skill, lift-required jobs so scheduling returns valid results, and when would you use Schedule Assistant versus an optimization run?
Read the worked answer
Model the job as a requirement group, not as two independent requirements. A requirement group lets you state that the job needs multiple resources with defined relationships - for example a hydraulic-skilled technician and an electrical-skilled technician, with a same-resource or crew constraint where the pair must work together, plus a facility requirement for the lift.
Set: duration and effort per requirement, the required skills and characteristics on each, territory, and a fulfillment preference that matches the promise you make to customers (closest, least busy, or a custom preference).
Tool choice:
- Schedule Assistant for interactive, one-job-at-a-time booking where the dispatcher wants to see candidates and choose. This is the everyday tool.
- Optimization for batch runs across many bookings where travel, time windows and priorities must be balanced - it will produce better overall routes than manual sequential booking.
Do not let dispatchers fall back to manual booking for the complex jobs; if the assistant returns nothing, that is a configuration signal, not a reason to type over it.
2. Schedule Assistant returns no results for the complex jobs. List the checks you would make, in a sensible order.
Read the worked answer
Work from the requirement outward:
- Requirement definition - are both skills present on enough resources, is the duration realistic, and is the time window achievable?
- Resource eligibility - do the technicians have the skills and characteristics recorded and current, are they enabled for scheduling, and are they members of the right pool or crew?
- Constraints - territory match, business closures and resource calendars, capacity already consumed, and whether the lift facility itself is bookable during the window.
- Settings - organisational time zone and working hours, and whether the fulfillment preference is so restrictive that no candidate satisfies it.
Then verify the pattern: if the failure is always at the pair level rather than the individual level, the same-resource or crew constraint is the likely culprit; a pair that is individually free but never simultaneously free will fail silently.
Fix the data first, then relax the preference to see whether candidates appear - that quickly separates a data problem from an over-constrained requirement.
3. Travel time is ignored and double bookings occur. How would you address both, and what would you report on?
Read the worked answer
Travel and double booking are two symptoms of the same gap: the schedule is built without a realistic timeline.
Address travel: enable travel time on the scheduling configuration so the engine allows for it between bookings, set the travel duration assumptions or use distance-based calculation, and ensure bookings carry start and end times that include travel. Without this, the engine will happily place two jobs that are physically impossible to attend.
Address double bookings: the manual bookings are the cause - a dispatcher booking by hand bypasses the capacity checks the engine would enforce. Stop the workaround by fixing why Schedule Assistant fails (requirement group constraints and resource data) rather than by policing the board. Use booking statuses to reflect travel, in progress and completed, and check the schedule board for technicians with overlapping bookings.
Report on: double bookings per week, on-time arrival rate against the promised window, travel time as a share of the working day, unbooked requirements older than 24 hours, and manual versus engine-generated bookings. That last ratio tells you whether dispatchers actually trust the tooling yet.