Razor Labs at Trust.AI Consortium: Pioneering AI Decision-Making Research
February 13, 2025
Artificial Intelligence is transforming industries, but understanding and trusting AI decisions remains a critical challenge. At Razor Labs, we are committed to leading innovation in this field. As part of the Trust.AI consortium, we set out to tackle one of AI’s biggest questions: How confident are AI models in their decisions, and can we trust them?
What is Trust.AI Consortium?
The Trust.AI consortium is led by the Israel Innovation Authority in collaboration with leading research institutions like Tel Aviv University, Ben-Gurion University, Bar-Ilan University, and Ariel University. The consortium was designed to explore Uncertainty Quantification—a crucial area of AI research that investigates how much confidence deep learning models have in their own predictions. This research has broad applications, from autonomous systems to industrial automation, ensuring AI models provide reliable and explainable outcomes.
Throughout this journey, Razor Labs collaborated with industry partners from the defense, telecommunications, and deep learning sectors, working together to push the boundaries of AI uncertainty research.
Razor Labs’ Contribution: Visual Inspection
As a leader in AI-driven industrial solutions, Razor Labs played a pivotal role in the Trust.AI consortium with its advancements in visual inspection—an AI-powered approach that enables automated video processing for real-time anomaly detection in manufacturing environments. By integrating uncertainty quantification, visual inspection enhances the reliability of AI-powered quality control, ensuring manufacturers can make data-driven decisions with confidence.
At the final Trust.AI consortium conference, the Razor Labs team presented our latest research on visual inspection, showcasing its impact on AI decision-making processes and uncertainty quantification. The conference, hosted by Microsoft, highlighted key findings, challenges, and future directions in AI uncertainty, reinforcing our commitment to pushing technological boundaries.
What’s Next?
While the Trust.AI consortium has officially concluded, we remain dedicated to driving AI innovation, bridging the gap between research and industry, and delivering cutting-edge visual inspection solutions to enhance industrial automation.
The visual inspection feature is already deployed as part of the DataMind AI predictive maintenance system to improve operational efficiency and increase throughput. This technology enables online monitoring of conveyor belts, ensuring the integrity of both conveyor belts and materials, as well as the entire crushing circuits and downstream machines.
