Visit scheduling
Visit windows, skills, and travel time.
Build a day plan that can actually run.
Model visit windows, required skills, travel, breaks, continuity, and urgent additions together, then create route candidates that people can inspect and adjust.
- Start from Excel
- Evaluate time windows, skills, and travel together
- Confirm before finalizing
Route plan
Candidate 02
Alex 09:00 Oak Home -> 10:35 Green TerraceBrooke 09:30 River House -> 11:20 North ClinicCasey 10:00 Treatment -> 13:15 Regular visit
Problem
The route-rebuilding work is heavier than the calendar entry work.
The hard part is not entering appointments. It is continuously rebuilding a feasible combination of visits, people, time, travel, and exceptions.
Visit windows overlap.
Some visits are morning-only, some must start after another task, and some require completion before a deadline.
- Travel before and after matters
- Waiting time also appears
Not every staff member can go.
Licenses, equipment, language, site knowledge, and customer preference can make a small pool of staff feasible.
- Qualified staff become the bottleneck
- Compatibility and continuity matter
Nearby on a map is not always feasible.
Straight-line distance is not enough when service time, parking, road time, base location, and direct travel matter.
- Travel squeezes the plan
- Each staff member may start elsewhere
One change ripples through the day.
Absence, cancellation, or emergency addition can force the whole remaining day to be rebuilt.
- Keep confirmed work as much as possible
- Notifications may also be needed
The planner is mentally solving this each time.
Visit windows×Staff conditions×Travel×Continuity×Changes×= runnable schedule
Mathematical model
Separate the rules before asking a solver to route the day.
Visit planning can be treated as a routing problem with time windows, staff skills, breaks, fixed appointments, and repair after changes.
01
Hard
Rules that must hold
- Visit time windowsEach visit starts and finishes inside the allowed service window.
- Required skillsOnly staff with the required skill, license, equipment, or language are considered.
- Staff availabilityAbsences, work hours, visit limits, and return-to-base rules stay visible.
02
Soft
Preferences to respect when possible
- ContinuityKeep the same responsible staff member when continuity has operational value.
- Preferred timePrefer a target time inside the allowed window when clients or sites request it.
- Workload balanceAvoid concentrating difficult visits, long travel, or urgent work on one person.
03
Objective
Metrics to improve
- Unassigned visitsReduce high-priority unassigned visits before optimizing convenience metrics.
- Travel timeMinimize total travel while keeping time windows and continuity explainable.
- Delay and overtimeShow when the available staff capacity cannot satisfy all demand.
Shortest distance alone does not make a good schedule.
We decide with you whether a few minutes of travel reduction is worth breaking continuity, or how many confirmed visits can move for an urgent case.
Field operations
Works for operations that match visits, people, and sequence.
The common question is who should visit which place, in which order, and at what time.
Home care and nursing
Handle visit windows, medical skills, continuity, weekly frequency, and absence repair.
Time windowLicenseContinuity
Field inspection and repair
Combine equipment type, parts, technician skills, emergency level, and travel from the base.
SkillPartsUrgent
Cleaning and maintenance
Represent entry windows, required equipment, multiple staff, and linked tasks.
Multi-personEquipmentEntry window
Sales visits and site surveys
Reflect priority, territory, relationship ownership, visit frequency, and direct travel.
PriorityAreaFrequency
Interactive example
Change conditions and inspect why the route candidate changes.
Change the scenario, absence, emergency visit, and rule strength to see how the candidate route changes.
Route map
Approximate routes
Coordinates are fictional relative positions. Production systems replace this with road-time matrices and verified address handling.
Review
Review points
Daily routes
Staff timeline
Unassigned
Visits that need a decision
Rule library
Model the operational rules that decide whether a route can be run.
The examples below are the kinds of rules to separate before choosing a solver or search method.
WIN-01HardVisit time window
Start and finish within each visit's allowed time window, or mark the visit as infeasible.
SKL-01HardRequired skill
Require the staff member to hold every skill, license, equipment category, or language needed by the visit.
AVL-01HardWorking availability
Respect staff absence, work start, work end, maximum visits, and fixed appointments.
BRK-01HardBreak window
Keep a meal break or rest window visible in the candidate route.
CON-01PreferenceContinuity
Prefer the previous responsible staff member when this does not break hard constraints.
CHG-01PreferenceMinimum change
Avoid changing confirmed visits when repairing the plan after absence or emergency additions.
BAL-01PreferenceWorkload balance
Balance visit count, service time, travel time, urgency, and difficult sites across staff.
TRV-01ObjectiveTravel reduction
Reduce total travel after feasibility, priority, and continuity have been considered.
UNS-01ObjectiveUnassigned explanation
Surface unassigned visits and the reason instead of forcing a misleading complete plan.
Explainable output
Output the reason for the plan, not just the plan itself.
A useful system does not hide shortage. It shows the candidate plan and the reason each exception needs human judgment.
Visit order and time
Show arrival, wait, start, finish, travel, break, and return timing for every staff member.
Unassigned reasons
Classify shortage reasons such as skills, time windows, visit limits, work hours, and fixed appointments.
Alternative comparison
Compare assigned count, travel, lateness, continuity, overtime, and change amount across alternatives.
Manual fixes and impact
Let people lock a confirmed visit or staff assignment, then recalculate only the remaining plan.
1Create candidatesCalculate from rules and priorities
→2Human reviewExceptions, relationships, field context
→3Lock part of the planKeep confirmed visits fixed
→4RecalculateRebuild the changed part
→From model to operations
Not just calculation: the system field teams actually use.
An optimization engine alone does not change operations. We design the screens and workflow around requests, customer data, staff schedules, maps, notifications, and results.
- Request intake and visit registerManage visit rules, address, service time, and priority.
- Staff mobile appCheck today's order, map, contact, arrival, and completion.
- Changes and notificationsReflect cancellations, absence, and urgent additions and notify the right people.
- Results and analysisAggregate gaps between plan and actuals, travel, delays, and utilization.
FIELDStaff smartphoneSchedule, map, report
OPERATIONSOperations screenRequests, edits, confirmation
MATHEMATICAL CORECalculation engineConstraints, routes, recalculation
DATACustomers, staff, resultsDB, API, audit log
Data
Start with the schedule files and rules you already use.
A first prototype can begin from one anonymized day or week. The goal is to confirm whether the rules can be represented and whether the candidate is useful.
Input
Data to review first
Visit list
Address or zone, time window, service minutes, required skill, priority, and previous staff.
Staff list
Work hours, base location, qualifications, visit limit, absence, and current location.
Business rules
Hard rules, preferences, priority order, break rules, fixed appointments, and change limits.
Current plan and results
Current plan, actual arrival and finish times, travel, cancellations, and manual changes.
Related experience
A system for checking visiting nurse schedules from a smartphone.
Finite Field has built systems that let care managers check and manage visiting nurse schedules from a smartphone. This page extends that operational base into candidate generation from multiple constraints.
View related workFit check
When custom development fits, and when it does not.
We first check whether the complexity and benefit justify a custom system.
+
Good fit for a custom model
- Many time windows, skills, and travel conditions overlap
- Planning depends on a specific coordinator's memory
- Absence or urgent additions force frequent rebuilding
- Packaged products still require heavy manual correction
- Schedules must connect to other operational systems
-
Consider packaged tools first
- The number of visits is small and fixed routes are enough
- Few time windows or qualification conditions exist
- A packaged scheduling service already covers the workflow
- Basic data or business rules are not yet organized
- The intake process should be organized before automation
Delivery process
Solve a small slice first, then connect it to field operations.
We separate calculation validation from system implementation so each phase can continue, change, or stop with evidence.
- 01DIAGNOSISFree check
Check the problem and data
Review current schedules, staff, visit rules, and change frequency.
- 02MODELINGDesign
Classify rules into three groups
Separate mandatory rules, preferences, and metrics, then agree how to score them.
- 03PROTOTYPEFixed scope
Run a small calculation prototype
Use one day or one week to generate candidates and compare them with the current plan.
- 04PRODUCTCustom quote
Implement the operations system
Implement intake, maps, manual editing, notifications, results, and permissions as one system.
- 05IMPROVEImprove
Improve with actual results
Use actual travel time, delays, and manual edits to refine the rules.
FAQ
Questions before automating visit scheduling.
These answers clarify the role of the page demo, map data, human review, and integration work.
Ask about your scheduleIs this page demo the production optimizer?
No. The page demo is a simple deterministic heuristic for explanation. Production work chooses a solver, search method, or hybrid approach after scale, rules, and response-time needs are known.
Can we start from Excel?
Yes. A first review can start from an anonymized visit list, staff list, rule memo, and current schedule.
Can the plan be repaired after an absence or emergency visit?
Yes. A repair mode can lock completed or already-notified visits, then recalculate only the remaining route and show what changed.
Can this connect to a map service?
Yes. Road-time APIs, distance matrices, and map displays can be integrated after cost, terms, caching, and privacy handling are confirmed.
Will the system confirm the final schedule automatically?
No. People should review unassigned visits, rule relaxations, and manual fixes before confirming the schedule.
Check your operation
Your visit schedule can be expressed as a calculable model.
Review your current Excel sheet or planning board and identify the data, rules, and first prototype scope.
- Anonymized data is acceptable
- We will say clearly if custom development is unnecessary
- No forced sales process