Hospital Inventory and Resource Integration System (HIRIS)

Inspiration

Inadequate inventory management practices lead to supply chain issues in healthcare (Link to article that explains these issues), which leads to medical resources not being available where needed. This is most acute in times of crisis such as pandemics or mass casualty events.

HIRIS's weighing system is inspired by weights in neural networks and coefficients in PID (Proportional-Derivative-Integral) controllers, both of which are architectures that can be customized to fit different needs.

What it does

HIRIS allows hospitals to use resources from other hospitals in times of need, by identifying and recommending nearby hospitals with a potential surplus of resources, based on the amount and type of resource needed.

How we built it

First, we identified the goals that the system of HIRIS needs to fulfil. From these goals, we determined a data structure that contains necessary data categories and is organized to streamline input, manipulation and retrieval. Then, we separated the system into multiple programs that interface with one another and the data structure, and establish milestones for their completion. Before coding each program, we also agreed upon the interfaces between the programs. While we worked on separate parts of the system, we performed frequent progress check-ups to test programs, , as well as identify unexpected new goals or changes to the interface.

Challenges we ran into

Lack of real data - created randomized mock data that has the same structure and purpose
Assumptions made for data categories - program designed to be adaptable for new data categories before release

Accomplishments that we're proud of

Completing our first hackathon
Researching and setting goals within the first two hours of the project
Distributing jobs according to the strengths of team members

What we learned

Integrating various programs by reading and writing to files, to create a program that is more organized and capable
Use of pandas Python module for data manipulation and retrieval

What's next for HIRIS

Improving performance of data retrieval algorithms, for scalability to larger datasets
Adding remote access capabilities
Integrating the weighing system with findings from system modelling studies, to provide more practical optimizations
Allow hospitals to specify base demand to retain a supply of resources if needed
Adding new/replacing current data categories to better suit medical industry needs
Improving the accuracy of Estimated Time to Arrival predictions
Improving security