Competitions

Developed an infrastructure that connects IoT devices with a web service through a Raspberry Pi gateway and custom APIs. A machine learning model is used to forecast the energy consumption of the building and to optimize the energy usage.

Data analysis of a environmental dataset, from the data understanding to outlier removal and statistical evaluations. Developed a forecasting model (LSTM) for time-series analysis and validated the process. We presented our solution at the Matrycs European Project launch event in Brussels, Belgium.

Evaluated the performance of different machine learning models for predicting the results of quality tests related to driveline mechanical components. Developed a desktop app to gather data from the sensors and provide the predictions to the operator in real-time.

Analyzed the leg of a humanoid robot to increase its degree of freedom and allow it to squat asynchronously. Designed and topology-optimized a prototype of the leg to be manufactured with additive manufacturing techniques.