Last Mile Route Planning for Parcel Distribution
Manual Route Planning faces many obstacles
Most professionals working with last mile route optimization for parcel distribution will attempt to have their resources do as many deliveries per vehicle per day as possible.
Obviously they will prefer to assign stops close to each other to the same vehicle. Many organizations have created a network of districts making it easy to decide to which route or vehicle a delivery should be assigned.
Both methods have a number of built-in challenges that makes it almost impossible to create exactly the routes that will efficiently utilize a fleet of vehicles and maximize the number of deliveries they can manage to do.
The planner who should decide which deliveries are close enough to each other to be
assigned to the same vehicles needs geographic insight at a very high level. The planning job tends to be even more complicated when vehicle capacities and work time are limiting factors and individual time windows for deliveries will make it increasingly harder to optimize last mile delivery without using dedicated route planning software.
Districts are not good for Last Mile Route Planning
Dividing a large area in to districts based on postal codes, may work for many organizations, but will not optimize daily operations. Accepting the fact that the best overall plan for a fleet of vehicles doing last mile deliveries is very much related to minimizing the drive time that vehicles spend driving from one delivery to the next it is easy to comprehend that 2 deliveries close to each other, but in 2 separate districts most likely would occur in less drive time if assigned to the same route.
Thousands of drive time calculations
When the route planning software Logistics Planner is
about to create an optimal plan for the delivery of, for example, 200 parcels using 4 vehicles the first step is to produce what is known as the OD (Origin Destination) Cost Matrix which means calculating the drive time from all 200 delivery addresses to the other 199 delivery addresses; 40.000 route plans with drive time that is. With this matrix the route planning solver can calculate total drive time for the deliveries no matter how the 200 parcels are assigned to the fleet of 4 vehicles. If we are about to visit 1.000 locations for parcel delivery the number of drive time calculations rises to 1.000.000.
The initial plan and how to improve it
Next step in optimizing last mile delivery for a fleet of vehicles is to create the initial plan that simply assigns tasks or parcels to the vehicles making sure that all restrictions and constraints are respected. For this initial plan the total costs are calculated as the total sum of drive time minutes and the cost of starting the necessary number of vehicles. Hereafter the software runs a number
of improvement steps. By moving tasks between vehicles and changing sequences billions of possible combinations are investigated in order to find the plan for the fleet that results in the lowest possible total costs. The improvement steps will often reduce total drive time for a fleet of delivery vans with 25 – 30 % – even when the fleet is as small as 4 vehicles. I addition to drive time and the number of vehicles Logistics Planner will also minimize idle time that may occur from arriving to a location prior to earliest time allowed for unloading.
Transportation of goods has many faces
Logistics Planner handles several specialties within parcel distribution an last mile optimization. One is distribution from a central warehouse facility, but the optimization software also handles scenarios where each vehicle has both pickups and deliveries along the route. Another scenario are with Online Supermarkets who need the option to optimize their last mile deliveries so every costumer gets his delivery within the exact hour decided by the costumer when placing his order.
More information and downloads
Please contact Leif Dahl Petersen for more information, pricing, documentation and trials
Berendsen Textil Service contracts a number of municipalities who offer laundry service to their elderly and disabled citizens. To do so Berendsen both pickup and deliver laundry to a large number of citizens every day. Agreed time windows must be respected and vehicle capacity together with the work time of the drivers must be taken into account when calculating the most effective route plans for the Berendsen fleet. This solution is delivered in cooperation with the Rapidis partner Pro-Sat.
Idémøbler, Ilva og Jysk; 3 co owned chains of 150 furniture shops deliver furniture to their costumers in Denmark and Sweden. Based on bookings in SAP or MS Dynamics AX a complete delivery plan for a fleet of 35 trucks is created making sure that the deliveries are both cost efficient and on time according to what is agreed with the receiving costumer. Integration and UI is delivered by the Rapidis partner PCSYS.
Washa, a commercial online laundry service both collects and delivers laundry at costumer locations which are both consumers and businesses. Daily routes are created and optimized, but Washa also uses Rapidis hosted services to dynamically calculate what would be the best and most cost efficient delivery option for a costumer who is about to order at the website; and the offer the costumer that option for a reduced fee.