Big VISION Initiative

Big VISION Initiative

posted November 5, 2025

Manufacturing companies often rely on visual inspection - looking for cracks, defects, or flaws - as a critical step to ensure product quality and safety. However, training today’s powerful artificial intelligence (AI) systems requires large collections of high-quality images. The Big VISION (Vision-based InduStrial InspectiON) Initiative, led by the University of Minnesota, addresses this need by creating one of the first large-scale, publicly available datasets for vision-based industrial inspection.

The goal of this project, led by MSI PI Shancong Mou (assistant professor, Industrial and Systems Engineering), is to gather a large image dataset, including products from diverse industries (e.g., electronics and steel manufacturing). Each image will be meticulously “annotated,” meaning trained inspectors and automated tools will identify and mark any defects at the pixel level. This will help AI systems recognize defects with greater accuracy.

Beyond collecting data, the team is pioneering new techniques to make the annotation process more efficient and cost-effective. Blending human feedback with AI-driven tools can accelerate labeling tasks that normally take countless hours. Finally, the Big VISION Initiative will bring together experts from academia and industry through workshops and forums. Sharing knowledge and addressing real-world challenges will enable power smarter manufacturing, reduce waste, and improve product quality worldwide. 

Professor Mou uses MSI for several projects to develop artificial intelligence/machine learning (AI/ML) techniques to enhance modeling and decision-making for complex systems. This research uses high-dimensional statistics, optimization research, machine learning, and computational science.

This project received a DSI (now DSAI) Data Sets Seed Grant and was featured during the Seed Grant Showcase in February 2025. This project falls under the Robotics and Sensors research area of the MnDRIVE initiative and the Foundational Data Sciences focus area. 

The DSAI Seed Grant program is currently paused and they are not accepting proposals. More information can be found on the Seed Grant website.

Discover Advanced Computing and Data Solutions at MSI

Our Services

MSI Related Topics

The Ag-Response Model
Research

posted September 24, 2025Nitrogen pollution from agriculture in Minnesota ground and surface waters is a major issue having negative effects on resident wellbeing and the environment.

Machine Learning Methods to Determine Arctic Methane Emissions
Research

posted September 10, 2025Methane has more than 80 times the warming power of carbon dioxide over the first 20 years after it reaches the atmosphere.

Carbon Emission Efficiency and Human Mobility
Research

posted May 7, 2025Efforts to mitigate global climate change necessitate the establishment of regional and local carbon emission control targets and reduction plans by countries, regions, cities, and companies.