J Intell Manuf
DOI 10.1007/s10845-015-1143-4
Probabilistic Boolean network modeling of an industrial machine
Pedro J. Rivera Torres
1
· E. I. Serrano Mercado
2
· Luis Anido Rifón
1
Received: 18 June 2015 / Accepted: 18 August 2015
© Springer Science+Business Media New York 2015
Abstract Theoretical modeling of manufacturing
processes assists the design of new systems for predic-
tions of future behavior, identifies improvement areas, and
evaluates changes to existing systems. A novel approach is
proposed to model industrial machines using probabilistic
Boolean networks (PBNs) to study the relationship between
machine components, their reliability and function. Once
a machine is modeled as a PBN, through identification of
regulatory nodes, predictors and selection probabilities, sim-
ulation and property verification are used to verify model
correctness and behavior. Using real machine data, model
parameters are estimated and a PBN is built to describe
the machine, and formulate valid predictions about proba-
bility of failure through time. Two models were established:
one with non-deterministic inputs (proposed), another with
components’ MTBFs inputs. Simulations were used to gen-
erate data required to conduct inferential statistical tests
to determine the level of correspondence between predic-
tions and real machine data. An ANOVA test shows no
difference between expected and observed values of the
two models (p value = 0.208). A two-sample T test demon-
strates the proposed model provides values closer to expected
values; consequently, it can model observable phenomena
(p value = 0.000). Simulations are used to generate data
required to conduct inferential statistical tests to determine
the level of correspondence between model prediction and
B Pedro J. Rivera Torres
privera@ece.uprm.edu
1
ETSET-Universidade de Vigo, Campus Universitario,
s/n, 36310 Vigo, Spain
2
Polytechnic University of Puerto Rico, 377 Ponce De Leon
Ave., San Juan, PR 00918, USA
real machine data. This research demonstrates that using
PBNs to model manufacturing systems provides a new mech-
anism for the study and prediction of their future behavior
at the design phase, assess future performance and identify
areas to improve design reliability and system resilience.
Keywords Bio-inspired modeling · Biological manufac-
turing systems · Probabilistic Boolean networks
Introduction
Probabilistic Boolean networks (PBN) are mathematical con-
structs that can be used to model Gene Regulatory Networks
(GRN). GRNs are collections of DNA segments within a cell
that interact indirectly with other segments and substances
in a cell in order to govern the expression levels of genes.
They can be used to better understand the general rules that
govern gene regulation in genomic DNA. PBNs are transi-
tion systems that satisfy the Markov Property (Markov 1954),
such that the probability that the system will take a transi-
tion from a given state to another depends exclusively on
the current state, and is not dependent on the past history
of the system. PBNs were proposed by Ilya Shmulevich in
several publications (Shmulevich et al. 2002a, b; Shmulevich
and Dougherty 2010) as an extension of Stuart Kauffman’s
Boolean Network (BN) concept (Kauffman 1969a, b). This
alternative to modeling GRNs combines the rule-based mod-
eling of Kauffman’s BNs with uncertainty principles. PBNs
consist of a group of constituent BNs that have assigned
selection probabilities, where each Boolean Network can be
considered a “context”. Data for the cells comes from dif-
ferent sources; each source represents a context of the cell
(Shmulevich and Dougherty 2010). In a given time t , a sys-
tem can be governed by one of the constituent BNs, and at
123