3 ms·
Yes we did, but most often you could just look at the delivery map and see what deliveries would go together. Optimization won't tell you which orders were not
by germansolo 6y ago
Yes we did, but most often you could just look at the delivery map and see what deliveries would go together. Optimization won't tell you which orders were not complete (waiting on an item that needed to be remade or was made wrong, etc.) and many times just eyeing the map and knowing the delivery area was all that was needed. Drivers also knew which shortcuts would be beneficial to use that the optimization would not know. The software at the time also would not be able to factor current traffic in the equation. It's been a while, maybe they do now, I don't know, I've left for greener pastures.
- scarface74 6y agoJust curious. I wrote software for the old ruggedized Windows Mobile devices for field service workers over a decade ago. The holy grail was route/cost optimizations. Near the end of my time there we integrated with a third party route optimization service.
- germansolo 6y agoThis was the main issue with the software we used. When it was slow, it would recommend that one driver take three orders to the same area while leaving others in the store with nothing, but when it was busy, it would suggest a driver only take one order even if there was three others in the immediate area and no drivers in the store. It didn't take into account the labor and mileage costs of its decisions. This was also a decade ago, I'm sure the software has been optimized for better results now.
- propogandist 6y agoI came across these guys sometime back on HN and have them bookmarked to try one day. Poking around the website, it looks pretty cool https://graphhopper.com/api/1/examples/#optimization https://graphhopper.com/api/1/examples/#optimization