Introduction

Overview

The Analysis Tools are a pre-built suite of advanced analytics capabilities within the Sight Machine platform, designed for in-depth exploration of time-series data.

These tools specialize in both univariate (single variable) and multivariate (multiple variables) analyses. Their primary function is to describe behavior over time, investigate correlations, and identify variables whose behavior might be impacting production processes or overall performance.

A foundational element of this suite is the underlying data preparation. All analyses are built on data that has been processed by Sight Machine’s AI Data Pipeline. This pipeline uses machine learning and AI to refine raw data by:

• Removing non-cycle variability, such as anomalies that occur during planned downtime.

• Modeling and summarizing parameter readings for each production cycle, which effectively eliminates noise from overly detailed, high-granularity readings, making the data optimal for exploratory analysis.

Correlation Tools

The following four tools on the Analysis tab, under Correlation, are focused on evaluating the relationship between manufacturing parameters, or correlation: 

• Correlation Heatmap
• Curve Fit Analysis
• Scatterplot Matrix
• Time-Series Correlation

Data Exploration Tools

The following seven tools on the Analysis tab, under Data Exploration, are focused on different ways to explore real-time contextualized data:

• Data Visualization
• Descriptive Statistics
• Event Timeline
• OEE Visibility
• Raw Data Monitoring
• Raw Data Visualization
• Timeline Analysis

Feature Benefits

• Allows users to track single parameter performance over time or uncover complex relationships and dependencies between multiple variables. 
• Provides a seamless, guided workflow for data exploration. By clicking on data points (e.g., in a Correlation Heatmap), the user is instantly presented with relevant subsequent analysis tools, which are automatically populated with the same initial parameters. This accelerates the investigation process. 
• Enables users to quickly uncover hidden relationships between variables and pinpoint the specific variables exhibiting behaviors that may be negatively impacting production. 
• Ensures that all analysis is performed on clean, reliable data. By removing non-cycle noise and variability, the pipeline prevents misleading results, allowing analysts to focus on true process impacts and correlations. 
• Simplifies complex data. By summarizing parameter readings per production cycle, the tools work with a noise-reduced dataset, which is ideally suited for fast, high-level exploratory analysis.  

Summary

The Sight Machine Analysis Tools offer a powerful, yet accessible, environment for advanced production data exploration. They provide the necessary capabilities—from simple variance tracking to complex multivariate correlation studies—to efficiently identify process-impacting variables and uncover deep relationships within time-series data. The true strength of this suite lies in its foundation: the AI Data Pipeline. By automatically cleaning and modeling the raw data, it ensures that all insights gained are based on highly reliable, noise-free information. The intuitive click-through functionality further enhances productivity by guiding users through logical, connected analysis paths, ultimately helping them track performance and rapidly solve production mysteries.