Transportation Operations

Road Transportation Network Visualization

Build Visualizations of your FTL Network Performance: deliveries/route, cost per ton and trucks size

Article originally published on Medium.

Following the series of Warehousing Operations Optimization, we will use the same methodology for improving Road Transportation efficiency by

Processing Data: extract unstructured transportation records and process them to build your optimization model
Improving Visibility: using Python visualization libraries to get clarity on current routing and truck loading rate
Simulating Scenarios: build a model to simulate multiple routing scenarios and estimate the impact on average cost per ton

💌 New articles straight in your inbox for free: Newsletter

I. Problem Statement

Problem Statement

Retail Stores Distribution with Full Truck Load (FTL)

1 Warehouse delivering stores by using three types of Trucks
(3.5T, 5T, 8T)
49 Stores delivered
12 Months of Historical Data with 10,000 Deliveries
7 days a week of Operations
23 Cities
84 Trucks in your fleet

Transportation Routes for 2 trucks covering 5 stores in 2 routes — (Image by Author)

Objective: Reduce the Cost per Ton

Method: Shipment Consolidation

In this scenario, you are using 3rd party carriers that charge full trucks per destination:

The table above shows rates applied by carriers for each city delivered for each type of truck. Observing costs per ton are lower for larger trucks, one lever of improvement is maximizing shipments consolidation when building routes.

Thus, the Route Transportation Planning Optimization's main target will be to cover a maximum number of stores per route.

II. Data Processing: Understand the Current Situation

1. Import Datasets

Before starting to think about the optimization model, your priority is to understand the current situation.

Starting with unstructured data coming from several sources, we’ll need to build a set of data frames to model our network and provide visibility on loading rate and list of stores delivered for each route.

Records of Deliveries per Store

Store Address

Transportation Costs

2. Listing of stores delivered by each route

Let us process the initial data frame to list all stores delivered for each route.

1 Route = 1 Truck ID + 1 Date

3. Add cities covered by each route

Let us now calculate Transportation Costs invoiced by carriers for each route:

Visualization: % Deliveries per Truck Size

(%) of Route per Truck Size (3.5T, 5T, 8T) — (Image by Author)

Impact of Average Truck Size (Ton) on Overall Cost per Ton (Rmb/Ton) — (Image by Author)

Insights

Average Truck Size: a large majority of small trucks
Cost per ton: the inverse proportion of cost per ton and average truck size

II. Understand Current Situation: Visualisation

1. Transportation Plan Visualisation

Objective: Get a simple visualisation of all deliveries per day with a focus on the number of different routes

Transportation Plan: January 2017 — (Image by Author)

Solution: Python’s Matplotlib grid function

Columns: 1 Column = 1 Store
Rows: 1 Row = 1 Day
Color = White: 0 delivery
Colors: 1 Color = 1 Route (1 Truck)

Visual Insights

Delivery Frequency: n deliveries per week
Number of Routes: how many different colours do you have a day?
Color = White: no delivery
Routes: do we have the same stores grouped from one day to another?

12 Months Overview

After optimization, this chart will help us to easily visualize the impact of new routing. A better routing means fewer routes per day so you’ll have fewer colours per line.

2. Geographical Visualization of Store Deliveries

Objective
Visualisation of geographical locations delivered in the same route

All Stores Locations (blue points) — (Image by Author)

Solution
OpenStreet Map + Matplotlib Scatter Plot

To get this visualisation, you can use the free tool OpenStreet Map. A detailed explanation can be found in this great article (Easy Steps To Plot Geographic Data on a Map, Link) written by Ahmed Qassim.

Visualization of the different routes covered per day

Visualisation of the different routes covered (1 colour = 1 route) — (Image by Author)

III. Next Steps

1. Measure the Environmental Impact

In addition to cost reduction, you can also target CO2 Emissions reductions by Optimizing your Transportation Network.

2. Routing Optimization: Number of Deliveries per Route

Dataframe with historical records processed
Current transportation plan
A model to calculate transportation cost per route based on cities delivered
Visualisation of the number of different routes per day
Visualisation of geographical locations delivered per Route

Next steps are

Routing: increase the number of stores delivered for each route
Fleet Allocation: ensure uniform workload distribution
Delivery Frequency: reduce the number of deliveries per week to increase the quantity per shipment
Simulate Impact: savings we can get from optimization listed above

About Me

Let’s connect on Linkedin and Twitter, I am a Supply Chain Engineer that is using data analytics to improve logistics operations and reduce costs.

References

[1] Ahmed Qassim, Easy Steps To Plot Geographic Data on a Map, Link

[2] Supply Chain Sustainability Reporting with Python, Samir Saci