A Service Production Planning Model Integrating Human Risk Factors

TitleA Service Production Planning Model Integrating Human Risk Factors
Publication TypeConference Paper
Year of Publication2014
AuthorsCao, NVi, Fragniere, E
EditorKersten, W, Blecker, T, Ringle, CM
Title of ProceedingsNext Generation Supply Chains
Volume18
Page345
Publisherepubli
Conference LocationHamburg
ISBN Number978-3-8442-9879-6; 978-3-7375-0340-2
ISSN NumberISSN (print) 2365-4430, ISSN (online) 2365-5070
Keywordshuman factor, production planning optimization, risk, service
Abstract

Most models of production planning based on mathematical programming tend
to assume constant technical coefficients. This assumption is realistic when the
production is based on machines as it is the case in manufacturing. On the
other hand, production planning in the service sector involves humans instead
of machines. Consequently, the assumption that all technical coefficient of the
mathematical program are constant cannot hold anymore. This is especially the
case for productivity parameters related to human activity. It is well known for
instance that in the service sector when administrative tasks are repetitive and
boring, working overload has a direct impact on the employee productivity.
We have adapted a manufacturing planning model producing industrial goods
into a service production planning model. In this service model, employees with
different job status (junior, senior and expert) are handling cases of specific
difficulties (simple, standard, personal and special). Then, we have introduced
a variable productivity formula into the mathematical program that takes into
account “plateau” levels assuming diminishing productivities. To do so the
mathematical program includes integer variables as well as non-linearity and
thus becomes a NLMIP (Non Linear Mixed Integer Program).
A fictitious case study is presented. The initial service production planning
model with constant technical coefficient leads to solutions involving job
specialization. On the other hand the model version with the variable
productivity formula offers a better workload balance and more possibilities of
job polyvalence reducing thus human risks such as burn-outs.