Niobia develops AI software to detect and diagnose defects in chip and battery manufacturing

AI for Defect Detection

Niobia, a startup based in San Francisco, has developed AI software designed to identify defects in the production of batteries and chips. The system aims to provide engineers with rapid insights into the causes of manufacturing failures, reducing the time typically spent on diagnosis from days to minutes.

How the Technology Works

The company's AI agents are deployed directly on manufacturing equipment, operating on self-contained GPUs. These agents function like sensors, capable of detecting minute flaws such as pinhole-sized scratches. Niobia's AI also centralizes production data, allowing engineers to interact with a chatbot to investigate issues and understand root causes.

Addressing Industry Challenges

The chip and battery manufacturing sectors frequently encounter defects during research and development, leading to significant financial losses and production delays. Each mistake can cost millions of dollars and extend manufacturing timelines. Niobia's solution aims to mitigate these issues by accelerating the defect resolution process.

Founders' Background and Market Focus

Niobia's co-founders, Gaurav Jha and Harveen Kaur, both hold PhDs in chemistry and have prior experience working at chip and battery manufacturers. This background informs their approach to addressing specific industry challenges. Jha noted that while the technology is specialized, it serves a multibillion-dollar market that includes major battery and semiconductor producers.

Technology Foundation

The AI agents utilize open-source models, which Niobia further trains with battery and chip research data. This training enables the system to identify small defects. The choice of open-source models was influenced by customer preferences to avoid sending sensitive company information to public AI clouds.