i
magine dispatching a driver to pick up a package in a city . One driver picking up one package is simple — just find the best route and go . However , when you have multiple packages , it becomes more complicated to plan the route . You have to consider a lot of variables , such as :
• Specific times when one package needs to be picked up or dropped off
• Special treatment for one piece because it ’ s coming from a strategic account
• Driver and vehicle constraints such as volume , weight , or timing .
These issues affect every delivery provider and are especially relevant to last-mile and courier service providers , who must handle complex routes most efficiently . The problem soon becomes huge when you have to factor in the need to provide multiple levels of Quality of Service ( QoS ) to premium accounts , ASAP delivery requests , meet SLA constraints and support multimodal delivery ( with different routes for bikes and cars , for example ).
It ’ s so huge that it may seem impossible . However , the team at Autofleet understands both the math and the reality of dispatch and routing planning .
Thanks to our optimized route planning software , we can build an adaptive solution that optimizes customer experience KPIs such as On Time Performance ( OTP ) and First Attempt Delivery Rates ( FADR ). It also helps improve fleet efficiency KPIs in deadhead distances and fleet utilization while considering improved driver experience .
Why is route optimization a complex problem ?
Optimized route planning is critical for a viable , sustainable business . It allows companies to reduce costs , increase efficiency , do more with fewer resources , and meet SLAs and customer expectations . However , route planning is hard because its complexity grows exponentially : The more stop points you add , the more possibilities there are . By a lot .
Take , for example , a plan with five stop points ( pick-ups and drop-offs ). This scenario has 120 different route combinations between these points ( A to B to C to D to E , A to C to D to E to B , etc .). With ten stop points , there are 3,628,800 optional routes . And if there were 20 , there would be more possible routes than grains of sand on a beach . That is , without factoring in multiple drivers and vehicles or timing constraints , which add to the complexity of the problem .
Even with the aid of software or a large team , route optimization is a difficult challenge . In fact , it is considered a classical NP-hard problem in computer science . Computers find it difficult to solve because it quickly grows too complex .
No algorithm guarantees a quick solution for all scenarios . In theory , you can “ brute force ” the
spring 2024 I customized logistics & delivery Magazine 23