

FluxGen is an IoT based startup which does water monitoring and management for industries. It aims at providing cost-effective Internet of Things (IoT) solutions to identify consumption pattern and reduce water expenditure.
Key Deliverables:
This is an IOT based startup working on the domain of sustainable renewable energy. The internship is related to develop a learning model to forecast the energy generation of solar panels using machine learning algorithms. Initially, the model predicts the solar radiation on which the energy is forecasted.
Forecasting of PV power output is a major concern to meet the increasing day to day electricity demands and also for operating a stable electric grid systems. PV modules, being the main part of solar power system, has many factors influencing its performance such as solar irradiance, wind speed, temperature, humidity and ambient temperature. This work proposes machine learning algorithms to forecast the energy generation of PV systems using the weather attributes. A day-ahead solar PV power output forecasting model is derived based on the historical power output data, weather forecasting data and advanced algorithms are applied to a PV station at IISc Bangalore. The accuracy of forecasting model applied on different algorithms is evaluated and compared by root-mean-square-error. This work has been published in ICSAC-2018.
Implemented the paper titled 'Personalised Travel Package with Multi-Point-Interest Recommendation Based on Crowdsourced User Footprints' after tailoring it for applications within the state of Karnataka.
Implemented an image classifier using Keras and Theano, trained on Imagenet dataset.Experimented with various configurations of nodes and layers in the neural network.