What is Causality?

The causality engine of Causal Activity Analysis ingests time vectored data from IIoT and IoT sensors, edge devices, database sources and enterprise systems.

To identify the actual causes that impact critical factors, such as loss of performance and production, plant downtime or poorly understood trends in retail sales figures, to pinpoint the factors affecting outcomes and makes possible the accurate prediction of likely faults, safety issues and performance or interrelationship between product sales trends in retail data.

To create visibility on these trend issues, in every enterprise operation from production, assembly, construction, energy resources operations, facilities management, logistics, healthcare services right through to supply chain and retail sales, delivering actionable insights and inform performance, productivity and safety in every enterprise…

CAA enables a more complete understanding of data to predict to a high degree of certainty:
  • What caused what and when
  • What will cause what and when
  • What could and should be done
  • To provide continuous improvement of processes

CAA makes decision-making more accurate by reducing risk of decisions being correlational only and not causal … with the corresponding likelihood of incorrect rectification of system problems due to wrongly detected or predicted causes.

Most data analytics use statistical correlation methods to try to identify the root cause of  critical factors such as unplanned downtime, safety issues or performance degradation. Examples from vendors in IoT:

  • Cisco (‘Attaining IoT Value’) …enable the company’s customers to perform real-time data correlation and, as a result, quickly react to irregularities.
  • Huawei (‘The IoT's Potential for Transformation’) …enables correlation-based process and productivity improvements.
  • Siemens PLM …quantitative statistical relationships to real-life usage, called customer correlation.
  • Industrial Internet Consortium …common issue in IIoT systems is correlating data between multiple sensors and process control states.

“ It has been well established that ‘correlation is not causation…”

Although proven to be of assistance analytically, correlational analytics is prone to producing spurious relationships and unreliable outcomes… e.g. between chicken and oil imports:

There is no empirical detection of causality, one does not have a causal impact on the other, let  alone provide the user with a measured causal coefficient of the actual, or the predicted, relational impact of cause on effect.

CAA’s objective is to provide the engineering team with a completely new, automated and  easily understood additional dimension to data analytics. One that reduces the risk of incorrect decision making and improves the certainty of analysis outcomes.

Acclaimed as a breakthrough technology, the impact of CAA’s causal analytics is profound and promises to greatly benefit the operations of asset-intensive industries where safety issues and productivity are paramount.