SMART SOFTWARE FOR CALCULATING OPTIMAL ROUTES, WHILE CONTINUOUSLY CHANGING SCHEDULES

PLANNING OPTIMIZATION

BRIEF

For one of Belgium’s leaders in the medical monitoring industry and External Occupational Health Service (EDPB), Mediaan Conclusion showed the benefits of adding smart technologies to their scheduling processes. Our customer handles over 300 requests every day in Belgium, resulting in 125 home visits for doctors. This requires an accurate and efficient route planning and appointment scheduling.

As a proof of concept, we developed a smart solution in just four (!) months. Thanks to our agile approach and our broad knowledge in Data Science, we automated their manual planning process. This could lead to a potential improvement of at least 40% in the current business processes and a yearly cost reduction of over 120.000 euro a year. For this project, we guided and supported our customer in numerous areas:

Developing a web portal
Optimizing route planning and appointment scheduling
Identifying the topic of interest using signalling functionalities
Visualizations – of geographic coverage
Data Science approach

THE CONCEPT

The idea was to develop multiple planning methods and evaluate their performance. We started by automating the current manual process as a reference point for improvement. Each planning approach creates a schedule with a list of accepted and rejected requests. The goal is to schedule as many requests as possible and minimize the total distance traveled over the route while adhering to the doctors’ availability time slots. The smart algorithms are decision support systems, meaning that the business users still have full control to choose the preferred alternative planning methods. Manual interventions are always possible throughout the whole scheduling processes.

CUSTOMER CHALLENGE

Home-visit requests come in continuously during the day while doctors are often constrained to specific time slots. To make it even more complex, there are multiple doctors in an area where multiple clients require home-visits, which shoud be assigned in a fair manner. The challenge was to make an optimal and efficient route planning and appointment scheduling for doctor visits. This process was done manually merely using a few basic predefined business rules. This caused human errors to occur due to the complex nature of the process.

MEDIAAN IN ACTION

To deliver the best possible solution, we developed a web portal and dashboards with the use of smart algorithms to project the right routes and create schedules. This smart solution was developed by our Data Scientists who identified the requirements and problem areas. The following techniques were used:

Angular
TypeScript
Bootstrap
Folium

Google maps
Python
Pandas
Microsoft Azure

SOLUTIONS IN DETAIL

A web portal– supporting the end-to-end process from importing requests for check-ups , planning and (re)scheduling, sending notifications to doctors, and data visualizations.
Route planning and appointment scheduling– supporting various algorithms from manual assignments to optimal and high performance approximation methods.
Signalling functionalities– identifying topics of interests and sending notifications to doctors and clients.
Visualizations– of geographic coverage of doctors, heatmaps, data distributions and outliers/anomalies in the data.
Data Science approach– From data collection, cleaning and structuring, to analyzing and visualizing the data, to trend reporting and predictions for ultimately supporting decision making.

RESULTS

The solution we developed resulted in an improvement of at least 40% in the current business processes, up to 60% when making small adjustments in the processes! Our automated algorithms show a reduction of 330.000 kilometres per year on traveling distance, with a yearly cost reduction of over 120.000 euro a year.

WANT TO SEE MORE?