• State Verified Answer
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 2 subscribers
  • Views 9 views
  • Users 0 members are here
Options
Related

Route optimization order logic — how does it decide?

Ray DeLuca

dispatch has been using the route optimization button and the order it spits out makes... questionable sense sometimes. like it'll have a tech drive 20 minutes north then back south past the starting point to hit a job that could've been first.

what's the actual logic here? is it just straight-line distance or does it factor in drive time, traffic, job priority, window times? the docs say "intelligent routing" but that's not exactly a technical spec.

asking because i've got techs grumbling about the sequences and i'd like to know if i'm defending an algorithm that actually makes sense or if i should just be drag-and-dropping everything manually.

  • 0

    Ray, happy to help clarify this! Route optimization in FieldPulse considers several factors in sequence:

    1. Job time windows — Hard constraints like "must arrive between 9-11am" take priority

    2. Technician skills and certifications — Won't assign plumbing jobs to an electrician, for example

    3. Drive time estimates — Based on historical averages, not live traffic

    4. Job priority — Emergency and high-priority jobs get weighted heavily

    5. Service duration — Longer jobs may be placed earlier to protect the schedule

    What it doesn't currently factor in: live traffic conditions, technician preferences, or customer "ease of access" (like gated communities that slow arrival).

    The "20 minutes north then back south" pattern you describe usually happens when a middle job has a tight time window that forces the sequence. If you want more manual control, you can pin specific jobs to time slots and let the optimizer fill around them.

    See Route Optimization: How It Works for the full breakdown.

  • 0

    Pro tip: if you're seeing weird backtracking, check whether someone set up recurring jobs with fixed time windows. Those act like anchors and can drag the whole route around them.

    We used to fight this constantly until we switched most recurring maintenance to "anytime during business hours" instead of specific slots. Route efficiency jumped maybe 15-20%.

    Also worth noting: the optimizer runs once when you hit the button. If a technician finishes early or gets delayed, the rest of that day's route doesn't automatically reshuffle. Some dispatchers will re-optimize midday if the morning goes sideways.

  • 0

    I'll add something Priya touched on but didn't emphasize: skills matching can really mess with perceived efficiency.

    If you have three techs and only one is certified for refrigeration work, that job might get placed oddly in the sequence because the optimizer is also trying to keep that tech's total drive time reasonable across their whole day. You're seeing one tech's route in isolation, but the algorithm is balancing all of them.

    We audit our skill assignments quarterly — found we had people tagged for things they hadn't done in years, which was constraining the optimizer unnecessarily.

  • 0 in reply to Gwen Albright

    that skills thing is interesting. we did a cert refresh last month and i don't think we updated profiles.

    still getting reports about that backtracking pattern though. might just pin the problematic jobs and call it good enough.

  • 0

    It's not just you. The optimizer is slow and the results are mediocre. We've been drag-and-drop for six months because the "intelligent" routing couldn't figure out that two jobs in the same industrial park should probably be consecutive.

    Save yourself the headache. Manual dispatch until they fix the traffic awareness.

  • 0 in reply to Mike Dolan

    Another way to think about this: the optimizer is solving a traveling salesman problem with time windows, which is computationally expensive. FieldPulse uses a heuristic that prioritizes schedule feasibility over pure distance minimization.

    That means it will accept a slightly longer drive to guarantee all time windows are hit, rather than risk a breach for marginal fuel savings. From a service level perspective, that tradeoff usually makes sense, even if individual routes look suboptimal.

    Our approach is to use optimization for the initial morning dispatch, then manually adjust based on real-time conditions. The hybrid method has worked well for a 40-tech operation.