During my internship at Ecolab, I had the opportunity to participate in a hackathon where my team and I leveraged Microsoft Azure and TensorFlow to develop a series of computer vision models. These models were trained on standard dishwashing loads to identify and classify objects within the frame.
The system captures images of dishwashing load and determines the optimal run time, water temperature, and soap quantity for the detected dishes. Varying treatment levels are required for different load types, which the model then classifies. This innovation was driven by Ecolab’s commitment to environmental sustainability, aiming to minimize water usage and ensure thorough cleaning for each unique load.
This project was my introduction to working with Microsoft Azure’s SQL databases, significantly enhancing my understanding of cloud storage systems.
We faced challenges in tuning the hyperparameters of our models and managing training times. Despite these obstacles, our team’s dedication and hard work paid off as we were awarded 2nd place out of 12 teams.
Reflecting on this experience, the project not only allowed me to deepen my technical expertise in machine learning and cloud databases but also demonstrated the impactful application of technology in promoting sustainability and efficiency in commercial settings.