Real-time understanding of your production environment
Factory Transform is a stream-processing data pipeline
Continuously generates a data foundation essential for both understanding and improvement of manufacturing operations.
Unique Challenges of Plant Data
- Manufacturing data was never generated for the purpose of analysis. It’s irregular, highly varied, out of order, and spread across multiple enterprise systems.
- Factory Transform uniquely addresses these challenges by integrating functions across the Sight Machine Pipeline, purpose built over a decade of working with streaming plant data, and designed to scale by applying advanced software development techniques to data processing.
Staging: Pick, Pack & Ship
A critical step for working at scale is staging: knowing what data you have, what it’s called, and how to attach to and model it. Sight Machine’s pipeline begins with tools for these tasks, and for linking meta-data to transforms and analysis
Stream and Transform
The pipeline is transparent, enabling visibility into raw data, transforms, and generated data tables. It’s configurable, and includes libraries of stateful transformations, as well as the ability to apply your own. And it’s robust. The pipeline accommodates late, missing, and out of order data, and it is built to handle changes in originating data environments.
Standardized Data Schemas and Common Data Models
Sight Machine’s Pipeline incorporates sophisticated data management tools: regions of data are identified and selected for analysis, derivative calculations and transformations are applied, and generated information is automatically produced as data is mapped. Data is streamed into Standardized Data Schemas, which in turn automatically generate Common Data Models. Common Data Models are applied to all manufacturing activities regardless of the product made or assets used.
These schemas and models have been proven across a wide variety of industries. Models include:
Understand production activity at different levels: machines, lines, and plants by modeling units of work done by machines. Types of work are limitless, but the idea of the unit of work is common, almost elemental. A unit of work is just the repeated cycle of activity by a machine.
Every time a machine performs a unit of work, a row of information is generated from all the data associated with that work. The unit of work is described with data from sensors on the machines, quality systems, MES, historians, ERP, and even ambient data like temperature and humidity, or other data about raw material characteristics.
- Units of work can be defined using signal based processing and or time boundaries
Understand everything that went into a unit of output along with its resulting quality. Sight Machine tracks the flow of material through the production process and associates all units of work to a unit of output.
- Units of output can be traced through the production process via serialization or a unique identifier available at each stage of production. In many cases, serialization is not available at each step so the combination of process values, line speed for example, and conditional time based offsets are used to associate units of work to each unit of output.
Sight Machine includes features for managing production data pipelines at scale
Preview, Copy, & Set as Production
- Quickly developing and testing streaming data pipelines is typically incredibly challenging and cumbersome.
- Test each transformation in the pipeline to ensure data integrity
- Copy production pipelines to test changes
- Set validated development pipelines as production seamlessness, with no downtime impact for analysts
Developer Workflow Integration
- Integrates with Git for version control
- Editable via a DAG or JSON
- Create your own transformation with Java
- Know if there is an issue in your production pipeline with built-in logs, alerting and notifications
Real Outcomes In Weeks
Getting started is easy
Curious about how we can help? Schedule a chat about your data and transformation needs.