From Pallet to Puppies: an insight from logistics to save animals

Project Type: Industry
LoDI Team: Monica Gentili, Erin Gerber, Kevin Gue, Stefanie Olga, Mark Cummins
Period: February 2016 – December 2016

Overview

The Kentucky Humane Society (KHS) is a private, nonprofit organization located in Louisville, KY. Founded in 1884, it is the state’s oldest animal welfare organization, and it is recognized as Kentucky’s largest pet adoption agency and largest no-kill animal shelter. Every year, thousands of animals are taken in, with over 50% coming from overcrowded shelters and the remainder being owner surrenders and strays. The number of lives saved defines the organization’s success.

Purpose and Need

In 2015 and 2016, KHS lives’ release rate was 96%. This includes animals that were adopted, returned to the owner, or (rarely) sent to rescues for specialized placement. Although KHS has had continued success, the organization wishes to better allocate adoptable animals to existing adoption facilities in order to free up space to accept more animals, and thus saving more lives. In 2012, KHS transitioned to admissions by appointment to ensure that they have adequate kennel space, eliminating euthanasia due to lack of space. This process has allowed more animals to be saved, but lack of space and inefficient utilization of space on the adoption floor are still among the most challenging issues that KHS has been facing.

Wanting to add a science-based approach, KHS asked researchers at the University of Louisville’s Logistics and Distribution Institute (LoDI), who normally deal with warehouse and distribution centers, to apply their research to puppies with the goal of maximizing adoptions and decreasing length of stay.

Methodology

In this project, LoDI team analyzed the shelter data; designed, implemented, and validated an optimization model; and implemented software with a user-friendly interface to facilitate the use of the mathematical model by the Kentucky Humane Society staff. The optimization model assigns animals to adoption locations so that the overall expected length of stay (LOS) across all KHS facilities is minimized, without exceeding the holding capacity of any of the facilities. The goal was to develop a mathematical model and deliver a software tool that could be used on a daily basis to better allocate animals to each adoption location and in order to minimize the expected length of stay.

The project was carried out through four main stages as follows:

Stage 1) Data Gathering and Analysis: Analysis of the data from online databases

Stage 2) Design, Implementation, and Validation of the Optimization Model: In this stage, we designed, developed, and implemented a deterministic optimization model to find the optimal solution to the animal allocation

Stage 3) User Interface Development and Implementation: We designed and implemented a user interface to facilitate the model applicability and provide an on-demand optimization tool.

Stage 4) Documentation, Training, and On-site Model Testing: The last stage of the project was devoted to (i) creating detailed instructions for using and maintaining the model; (ii) conducting training to hand off the finished model to KHS users; and (iii) testing the model by using real data for at least a six month period.

Results and Impacts

The initial results show that the allocation model provides a substantial improvement compared to the traditional KHS allocation. This outcome not only verified that the allocation model worked properly, but it also assured that this will allow KHS to make more efficient use of the limited space they have. The allocation model created for the KHS is a good start in improving their adoption process. Overall, employees at KHS were satisfied with the developed tool which makes their decision-making process faster and more effective.