TL;DR
- Multi-stop route planning sequences a single tech’s whole day into one optimized route — whole-day optimization, not per-job assignment.
- Manual sequencing achieves only 60-70% efficiency past four or five stops. Algorithmic optimization consistently hits 92-97%.
- Ten jobs create 3.6 million possible sequences. Fifteen create over 1.3 trillion. No human can scan that search space in real time.
- Dependencies (diagnostic-before-repair), time windows, skills, and equipment are all hard constraints the solver respects simultaneously.
- Anytime scheduling is the single biggest lever. Routes with 80% anytime jobs run 58% more drive-time efficient than routes with all fixed windows.
If a tech has eight stops on Tuesday, there are 40,320 ways to order them. Twelve stops? 479 million. Fifteen? Over 1.3 trillion. Your senior dispatcher is not picking the best one — they are picking a “good enough” one in about four minutes and moving on to the next phone call. Multi-stop route planning is the layer of AI dispatching that scans the whole search space and returns the actual best route, every time, in under two seconds.
This guide explains what multi-stop route planning actually does, why manual sequencing breaks past four or five stops, how the solver handles dependencies and time windows, four real sequencing scenarios from production deployments, when multi-stop optimization stops helping, and how to run it inside FieldCamp’s AI dispatch software. The numbers below come from 50+ shops running multi-stop sequencing in HVAC, plumbing, lawn care, pool service, and appliance repair — same job list as before, 30-40% less drive time.
Why Multi-Stop Routing Breaks Manual Dispatching
Direct answer: Multi-stop routing overwhelms human dispatchers because the search space grows factorially. Beyond four or five stops, manual sequencing achieves only 60-70% route efficiency compared to algorithmic optimization, which consistently delivers 92-97%. The hidden cost is not visible in the morning plan — it shows up as backtracking, missed windows, wasted fuel, and burnout from windshield time.
Faced with this complexity, dispatchers fall back on shortcuts:
- “Send whoever’s closest.”
- “Give Mike all north-side jobs.”
- “Drop emergencies on whoever frees up first.”
These rules work for three or four stops and break at eight or more. An HVAC company with four technicians each handling 8-10 daily tune-ups typically loses 8-12 hours of weekly labor to manual sequencing alone. The fix lives in proper AI route optimization driving the day, not in better whiteboards.
What the Optimizer Actually Does
Direct answer: Multi-stop planning is the operator-side instance of the Vehicle Routing Problem with Time Windows. Where pure route optimization asks “what is the shortest route that visits all stops once?”, multi-stop planning answers it for one tech’s actual day — balancing drive time, time windows, skills, equipment, dependencies, and the 15-minute travel-time ceiling between stops all at once.
Manual dispatching can satisfy two or three of these constraints at once. The optimizer handles all of them simultaneously, anchored in the 50-variable decision model that drives every AI dispatch scheduling decision. Capacity constraints follow the capacitated vehicle routing playbook — see capacitated vehicle routing for the depth.
Sequencing — The Core Operation
Direct answer: Sequencing decides the order of stops on a single tech’s route. It is distinct from job assignment, which decides who does the job. Assignment happens first; sequencing turns assigned jobs into a drivable route. Single-stop thinking fails because each individual choice looks rational, but the cumulative path builds in backtracking and missed windows.
A worked example: a dispatcher assigns Job 1 to the closest tech (2 miles away), Job 2 to the next-closest (3 miles), and so on. By Job 8, the route has three instances of backtracking and 1.2 hours of unnecessary drive time. A holistic re-sequence on the same eight jobs produces a route with zero backtracking and 2.3 hours total drive time — a 44% improvement.
| Approach | Jobs | Drive Time |
|---|---|---|
| Single-stop thinking | 8 | 4.1 hours |
| Multi-stop optimization | 8 | 2.3 hours |
Sequencing is where job dependencies are enforced. If a diagnostic must precede a repair, the optimizer will never place the repair stop first, even if doing so would shave drive time. Dependencies are hard constraints; drive time is a soft constraint. The biggest single lever for sequencing efficiency is anytime scheduling — date-specific but time-flexible jobs let the solver cluster geographically rather than chase the clock. Routes with 80% anytime jobs are roughly 58% more drive-time efficient than routes with all fixed windows.
See Multi-Stop Sequencing Live
Bring your tech roster and a typical day’s jobs. We will overlay the optimized sequence on your real route map and show the recovered hours.
Real Sequencing Scenarios From Production
Direct answer: Four representative scenarios show how multi-stop planning behaves under different constraint mixes — geographic, fixed-window, emergency insertion, and dependency-driven. Each one is a real pattern we see in shops running AI job scheduling against live job mixes.
Pool Service Circuit
Twelve weekly maintenance stops, geographically clustered, all anytime. The solver builds a single efficient circuit through the service area, takes total route time from 5.8 hours to 3.2 hours, and produces zero backtracking — a 45% reduction in drive time.
HVAC Maintenance Day
Eight seasonal tune-ups, EPA-certified tech required, three jobs with fixed two-hour windows and five anytime. The solver respects the three fixed windows and clusters the anytime jobs around them, hitting 2.1 hours of drive time and zero window misses.
Plumbing Emergency Insertion
A six-stop route is mid-execution when an urgent call arrives at 2:17 PM. The solver inserts the emergency as stop #4 and re-sequences the remaining stops to minimize disruption — total schedule delay of 12 minutes versus 47 minutes for a manual append-to-end approach. Insertion completes in 200-500 ms. The full lifecycle is owned by dynamic rerouting.
Appliance Repair Circuit With Dependencies
Five stops with parts-dependent sequencing — diagnostic must happen before repair, parts pickup between them. The solver produces: diagnostic → parts pickup → repair → remaining stops. Drive time drops from 2.3 hours to 1.6 hours and the dependency chain is never violated.
Hard and Soft Constraints — What the Solver Will Not Bend
Direct answer: Hard constraints are non-negotiable; the solver will never produce a route that violates one. Soft constraints are weighted penalties; the solver minimizes them but may accept a small violation to satisfy a hard constraint or improve total cost. The split is the reason multi-stop optimization produces drivable routes instead of mathematically clean but operationally broken ones.
| Constraint type | Examples | How solver treats it |
|---|---|---|
| Hard | Skill certifications, dependencies, equipment availability, CONFIRMED job status | Never violated |
| Soft | Drive time, fairness, customer preference, end-of-day location | Minimized via penalty score |
| Configurable | 15-minute travel ceiling, overtime cap, lunch window | Hard by default, can be relaxed |
This is the same constraint model that powers AI Command Center across the rest of the day. If you want to size what changes when you flip a configurable constraint, run scenarios in FieldCamp’s Route Optimization screen before committing.
KEY TAKEAWAY
Hard constraints (skills, dependencies, equipment) keep the solver from producing impossible routes. Soft constraints (drive time, fairness) keep it from producing ugly ones. Together they’re why algorithmic sequencing produces routes that actually run.
When Multi-Stop Optimization Stops Helping
Direct answer: Sequencing delivers diminishing returns under three conditions: when 80%+ of jobs have narrow fixed windows, when jobs spread across 100+ miles, or when techs run three or fewer stops per day. The honest answer in those cases is that the solver still helps, but the gain may not be worth the operational complexity.
- All jobs have narrow windows. When 80%+ of the day is locked into fixed two-hour windows, optimization savings drop to 10-15%. The clock dictates the sequence and geographic clustering cannot help. Loosening windows to anytime where possible unlocks the full lift.
- Jobs spread across 100+ miles. Geographic clustering does not help when stops are too far apart to cluster. The fix is chaining work across days through multi-day scheduling, which absorbs the geography problem at a higher level.
- Three or fewer stops per tech. With six possible sequences, a dispatcher can eyeball the best one and the optimization overhead is not worth it.
Manual override is the right answer when the dispatcher knows something the algorithm does not: a VIP request, an emergency that overrides optimal sequence, a long-standing tech-customer relationship, or weather affecting a specific area. AI handles the heavy lifting; dispatchers handle the exceptions.
See AI Dispatcher In Action
Watch FieldCamp re-sequence a typical 10-stop day in under two seconds and walk through the before/after side-by-side with our team.
Why Anytime Scheduling Unlocks the Biggest Gains
Direct answer: Anytime jobs are date-specific but time-flexible — “service this property on Tuesday, any time during business hours.” That single relaxation lets the solver cluster geographically instead of chasing the clock, typically unlocking 40-60% better drive-time efficiency than routes with all fixed two-hour windows. The hard part is convincing customers to accept anytime windows.
The customer-facing trick is offering tighter ETA notifications in exchange for a wider window. “We will arrive between 8 AM and 12 PM, and we will text you a 30-minute heads-up when the tech is en route” is a fair trade most customers accept. The result is better routing on your side and a better experience on theirs — fewer “where are you” calls because they know exactly when to expect the truck. Pair this with a clean work order management flow so the tech has full job context the moment they arrive.
PRO TIP
The fastest way to lift route efficiency 40-60% isn’t a better solver — it’s converting more jobs from fixed two-hour windows to anytime. Train your CSRs to offer wide windows with proactive ETA texts as the default.
Multi-Stop Planning in FieldCamp
Direct answer: FieldCamp’s Route Optimization screen runs the proprietary deep-optimization engine against a job list, balancing drive time, window compliance, dependencies, skill matching, and workload fairness simultaneously. Two modes are exposed: single-tech (perfect one route) and multi-tech rebalancing (redistribute jobs across the team).
The dispatcher workflow is short: create jobs, mark them anytime where possible, navigate to Team Tracking → Route Optimization, select tech and date range, click Optimize, review the before/after comparison, confirm. Measured across 50 dispatcher sessions, total workflow time averages 1 minute 47 seconds. Customers report 30-40% weekly drive-time reductions after switching from manual planning, with no increase in missed time windows. Use the labor cost calculator to size what those recovered hours mean for revenue.
To get this running on your account, follow submitting a job for AI dispatch and the AI job scheduling quick guide. The bigger picture lives in the AI route optimization explained playbook.
Run A Smarter Dispatch Day
Sequence 10-15 daily stops in under two seconds. Book a 30-minute walkthrough and we will run your actual job list through the optimizer.
Frequently Asked Questions
How many stops can AI optimize in a single route?
Routes with 50+ stops are well within range, though most field service routes contain 6-15 stops per technician per day. Beyond four or five stops, manual sequencing already underperforms algorithmic optimization. The solver scales linearly with stop count — adding stops does not blow up runtime the way the factorial search space suggests.
What is the difference between route optimization and job assignment?
Job assignment answers “who should do this job?” — picking the right tech based on skills, location, and workload. Route optimization answers “in what order should this tech complete all their jobs?” Assignment happens first; sequencing turns the assigned jobs into a drivable route. Multi-stop planning is the sequencing layer.
Why does anytime scheduling improve route efficiency?
Anytime jobs are date-specific but time-flexible. That single relaxation lets the solver cluster geographically instead of chasing the clock, typically unlocking 40-60% better drive-time efficiency. Routes with 80% anytime jobs run 58% more efficient than routes with all fixed two-hour windows.
Can AI re-optimize routes when emergencies arrive mid-day?
Yes. The emergency is inserted at the optimal position and remaining stops are re-sequenced to minimize disruption. Insertion completes in 200-500 ms. The full lifecycle is owned by dynamic rerouting, which handles GPS deviation, status updates, and external triggers along with emergency insertion.
How does AI handle job dependencies?
Dependencies are treated as hard constraints. A repair will never be sequenced before its diagnostic, even if drive time could be reduced by doing so. The solver enforces the dependency graph before considering drive-time optimization, which means dependent stops always run in the correct order regardless of geography.
Can FieldCamp optimize recurring service routes?
Yes. Recurring services like lawn care, pool cleaning, and pest control are clustered geographically into weekly or monthly routes that minimize mileage and keep technician schedules predictable. The solver respects recurring service-day preferences while clustering route geography for maximum efficiency.
What happens if I disagree with the optimized sequence?
You override. The dispatcher view shows the optimized sequence side-by-side with the original, with a per-stop reasoning summary. One click to accept, override, or request an alternative. Manual overrides feed back into the system so future optimizations weight the dispatcher’s preference appropriately.
Continue reading
- How AI reduces drive time — six mechanisms compressing windshield time across the day.
- Dynamic rerouting — live sequencing updates when reality bends the morning plan.
- AI route optimization explained — the framework that turns multi-stop planning into a daily routine.
- Capacitated vehicle routing — capacity constraints layered on top of sequencing.