WEATHERAI
- /
- Success stories
- /
- WEATHERAI
WaterView uses advanced computer vision and AI technologies to transform common cameras into multivariable sensors for real-time granular weather and environmental monitoring, reducing costs. First-generation models require expensive hardware and software and device calibration, while WeatherAI’s second-generation AI analysis is more portable and scalable and trained on optimized datasets. This could be a significant breakthrough in improving the ability to monitor the environment and climate on a global scale.
Start date: 01/06/2021
Duration in months: 12
Problem Description
WeatherAI wants to develop a second generation of AI algorithms that is more portable and scalable and trained on optimized datasets.
Goals
New services
Challenges
Parallelize training algorithm on HPC to achieve the required accuracy in inference.
Innovation results
GPUs were used for the model training. Parallel CPUs were used for image augmentation. Hyperparameters and datasets were optimized for 5 models.
Business impact
Test-before-invest. It is possible to test the feasibility of using AI to develop weather monitoring analytics, before buying the hardware