Kurt DeMaagd, Chief AI Officer at Sight Machine, discusses democratizing industrial data with generative AI and how their company helps manufacturers extract and transform data into useful formats. They also cover the architecture of Sight Machine's platform, their support for edge devices and cloud computing, industries and customers they work with, and the capabilities of their AI interface including multilingual support.
Factory Co-Pilot empowers individuals to interact with manufacturing data in a user-friendly way through generative AI and a chat-based interface.
Site Machine's data-first approach enables users to access and blend real-time data from different manufacturing sources, supporting both cloud and edge computing.
Factory Co-Pilot aims to expand the user base by providing a more intuitive and user-friendly interface, allowing individuals from various roles and backgrounds to effectively leverage manufacturing data.
Deep dives
Democratizing Industrial Data with Generative AI
Site Machine introduces Factory Co-Pilot, a solution that aims to bridge the gap between traditional data users and those who may not have the expertise. Factory Co-Pilot leverages generative AI and a chat-based interface to make data more accessible and intuitive for all users. By enabling users to ask questions and receive answers in a conversational manner, Factory Co-Pilot empowers individuals to interact with and derive insights from manufacturing data in a user-friendly way.
Data First Approach and Connectivity
Site Machine takes a data-first approach, focusing on extracting and transforming data from various manufacturing data sources. With a robust set of connectors, including support for OPC UA, Site Machine enables users to access and blend real-time data from different plant floor systems, historians, quality management systems, and more. The platform supports both cloud and edge computing to cater to different requirements, providing flexibility and accessibility for data analysis and insights.
Wide Range of Industries and Users
Site Machine caters to a wide range of industries, including discrete, process, and consumer packaged goods. Typical users of the platform are process engineers, quality engineers, and other data-savvy individuals on the plant floor. However, Factory Co-Pilot aims to expand the user base by providing a more intuitive and user-friendly interface, allowing individuals from various roles and backgrounds to effectively leverage manufacturing data for problem-solving and decision-making purposes.
The Manufacturing Data Platform
The manufacturing data platform provided by the company aims to bring together various data sources in real time, blending information from quality management systems, the plant floor, and ERP systems. This integrated approach enables users to analyze the entire system and solve complex problems. It also eliminates the need for manual data gathering and allows for real-time access to information. By prioritizing a data-first approach, the platform facilitates faster application development and accelerates time to value.
Application of AI and Machine Learning
The company leverages AI and machine learning in various ways within their manufacturing data platform. They use AI algorithms to automatically detect, organize, and clean large volumes of diverse data, ensuring high-quality output. At the application layer, they offer out-of-the-box tools like recommendation systems, such as their Cookbooks application, which optimizes machine settings based on real-time data. They also provide customized analytics for specific use cases, like scheduling optimization in a large dairy processor. The platform's flexibility allows for the selection of the best suited AI approach for each use case to deliver value to operators, engineers, and other users.
Peter Seeberg talked to Kurt DeMaagd, Chief AI Officer at Sight Machine - the data platform BASF, Nvidia, Komatsu, Heineken, Nike, Nissan and Jeep are using.
Thanks for listening. We welcome suggestions for topics, criticism and a few stars on Apple, Spotify and Co.