Causal AI

Causal Artificial Intelligence distinguishes between correlation and causation. Correlation is not Causation. It is not necessary that everyone who smokes will get cancer. But that is what the present day AI modules may predict. Correlation based AI application make a prediction based on ‘what’ has happened and not ‘why’ it has happened. 

If we do not know the root cause, we could easily make poor decisions and support ineffective and prejudicial policies and judgements. Root cause analysis has to do with causation: with providing an answer to the question why an incident happened, rather than to the question: what, where, and how much.

Causal models are mathematical models representing causal relationships within an individual system or population. They facilitate inferences about causal relationships from statistical data: relationship between causation and probability.

FMEA and FTA establish the relationship between failure cause and its effect. The probabilistic relation between cause and effect is determined through Critical Analysis (CA). Bayesian Network (BN) is a graphical model representing the joint probability distribution P(X) on a set of random variables X = {X1, …., Xn} defining probabilities P ∈ [0,1] for each possible state (x1, , xn) ∈ X1,dom, …,Xn,dom where Xi,dom is the domain of definition of each variable.

FMECA/FTA are used primarily for the reliability of system operation or equipment to keep it always functional. While BN is used to diagnose a system when it is facing failure and also to predict it in order to avoid its appearance. We have transformed an FMECA/FTA to a BN. The structure of BN is derived from the effects, causes and failure modes present in FMECA/FTA.

In this BN, we utilise following two scenarios. In the first, we suppose that there is a failure and we want to identify its causes. Indeed, the diagnosis makes it possible to explain the causes of failures by the back chaining in the BN. In the second scenario, we suppose there’s not a failure in the equipment, but we hope to know the probability of the consequences given the causes of failures is known in advance. So, we predict that by the front chaining in the BN.

We at Reliability Centre India are linking the FMECA/FTA to BN.