Lumafield
Lumafield builds industrial CT scanners and the software to make sense of the data they produce. Founded in 2019, the company targets a specific problem in manufacturing: the inability to see inside a finished part without destroying it. Their Neptune scanner is designed to be accessible - cheaper, smaller, less complex than legacy industrial CT systems - giving engineers non-destructive internal inspection capabilities that were previously reserved for aerospace budgets. The security surface here isn't traditional network defense, but it's real. These are internet-connected industrial hardware systems shipping CT scan data to a cloud-based platform. Voyager, their browser-based analysis tool, handles sensitive manufacturing data - internal geometries, defect maps, quality-control records - across distributed teams. That means protecting proprietary design IP in transit and at rest, securing device firmware against tampering, and managing access control on data that could reveal a manufacturer's process weaknesses or product flaws. The technical stack spans computer vision , AI-powered analysis , industrial IoT , and web platforms . Anyone working security here needs fluency in both the physical attack surface - embedded systems, scanner firmware, data acquisition pipelines - and the cloud layer where analysis happens. The threat model includes supply-chain compromise of industrial hardware, exfiltration of proprietary scan data, and adversarial manipulation of AI models used for defect detection.
Lumafield builds industrial CT scanners and the software to make sense of the data they produce. Founded in 2019, the company targets a specific problem in manufacturing: the inability to see inside a finished part without destroying it. Their Neptune scanner is designed to be accessible - cheaper, smaller, less complex than legacy industrial CT systems - giving engineers non-destructive internal inspection capabilities that were previously reserved for aerospace budgets.
The security surface here isn't traditional network defense, but it's real. These are internet-connected industrial hardware systems shipping CT scan data to a cloud-based platform. Voyager, their browser-based analysis tool, handles sensitive manufacturing data - internal geometries, defect maps, quality-control records - across distributed teams. That means protecting proprietary design IP in transit and at rest, securing device firmware against tampering, and managing access control on data that could reveal a manufacturer's process weaknesses or product flaws.
The technical stack spans computer vision, AI-powered analysis, industrial IoT, and web platforms. Anyone working security here needs fluency in both the physical attack surface - embedded systems, scanner firmware, data acquisition pipelines - and the cloud layer where analysis happens. The threat model includes supply-chain compromise of industrial hardware, exfiltration of proprietary scan data, and adversarial manipulation of AI models used for defect detection.