Milk procurement route optimisation using GIS

Milk procurement route optimisation using GIS

SHARE

Darsh Worah
Dy. Manager, SAS Group
[email protected]

Mena Paghadar
Dy. Manager, SAS Group
[email protected]

Subir Mitra
Sr. Manager, SAS Group
[email protected]

National Dairy Development Board, Anand, India

Abstract
Using GIS, it is possible to depict milk procurement routes on digitised maps and the process of visualising alternative milk routes can be made easier as compared to the current manual processes. It is accepted world over that vehicle route planning is a difficult combinatorial optimisation problem which needs to be addressed in supply chain management, especially when the raw material has to be sourced from geographically dispersed areas.

Introduction
The quantity of milk procured from the villages by any dairy organisation is one of the most critical factors in its overall planning of activities. Milk procurement activities are spread over large geographic areas and involve a large number of far flung village level dairy societies and thousands of farmers who are members of these societies. Milk procured from villages has to be transported to a chilling station or bulk milk cooler for temporary storage and then further to a dairy for processing and packaging.

The milk procurement routes of the dairy organisation are normally planned with the twin objectives of maximising the volume of milk procured/km and minimising the cost of milk procurement/litre, under the given set of its current constraints. Further, seasonal or economic constraints may lead to variation in these milk routes.

For this case study, 138 village level societies of a typical farmers’ dairy organization in Maharashtra were considered, which procured milk at a particular point in time and delivered it to the nearest dairy plant in the city.

The purpose of the study was to explore the possibilities of introducing GIS in a farmers’ organisation for milk procurement route optimisation and to understand the practical problems which are likely to be encountered in introducing the latest technology in such an environment.

Methodology: Flow chart

Description of methodology

  1. Locations of dairy societies and tracks of existing milk routes were collected using hand-held GPS device. Village level milk route supervisors were trained to use the hand held device “on-the-fly” for 2-3 days by moving on the milk procurement vehicle and saving routes and milk pickup points on those routes. The locations of dairy societies and the existing routes were then plotted on map (Fig.1).
  2. Training was also imparted to the MIS staff of the organisation so that they could manage the information on their own. Thus, they could send the collected raw data to us for final processing.
  3. Updated road network was modelled in the Network Analyst.
  4. Vehicle routing utility of ArcGIS Network Analyst Extension was used to optimise the routes.
  5. Details about vehicle type, carrying capacity, maximum travel distance and maximum orders were given as input (Fig 2).
  6. Buffer of 1500 metres was given to locate the pickup points along the road network.


Fig 1. Existing Milk collection routes


Fig 2. Inputs for Vehicle routing problem.

Results
The total distance covered by existing milk routes for procuring 10,800 kg/shift was 1445 km and there were 13 routes with various capacities of vehicles (Table 1). After optimisation, the total distance could be reduced to 1351 km with 11 routes (Table 2). Savings in transportation cost per shift was Rs. 504. We also observed that the optimised routes engage different vehicle type and visit different collection points compared to existing routes (Fig 4).

Challenges in implementation

  1. The availability of an updated village level road network is a major bottleneck as it is a crucial input for the route optimisation software. Though our intention was to use the entire possible road network for optimisation, the exercise was finally done only for the road networks as could be captured by the village level supervisors.
  2. Farmers’ organisations have more pronounced challenges in recruiting and retaining of employees with IT background. However, collecting field level through GPS is not a difficult task, as the existing village level supervisors can use the handheld GPS device easily.
  3. Utilising the route optimisation software by itself is a small part of the problem, the bigger problem is organisation of a proper database and keeping it updated for analysis in real time.


Table 1. Existing route details


Table 2. Optimised routes – details


Fig 3. Optimised Milk collection Routes.


Fig 4. Existing and Optimised route example

Conclusion
It was observed that implementing GIS technology in a farmers’ organisations would be quite beneficial, given the current scenario of availability of cost effective hardware and software and also given that proper support and training is provided. Initial hand holding with the manpower and deployment of user friendly software is also required.

The main learning from this case study was that GIS technology can be effectively used by farmers’ organisations to validate the efficacy and design of their existing milk collection routes and also to discover the scope for improvement.

Thus, using GIS technology for milk procurement route optimisation would contribute significantly to these dairy organisations in terms of saving money and time along with its concomitant reduction in CO2 emissions.