Content uploaded by Emile Glorieux
Author content
All content in this area was uploaded by Emile Glorieux on Jun 06, 2022
Content may be subject to copyright.
VIRTUAL ENGINEERING TO DESIGN ADVANCED MANUFACTURING SYSTEMS
Emile Glorieux
Terie Purse
Business & Factory Optimisation
Digital Engineering Group
The Manufacturing Technology Centre (MTC)
Ansty Park, Coventry CV7 9JU, (UK)
ABSTRACT
Designing advanced manufacturing systems (AMS) is a complex task needing to consider the re-
configurability requirements, how these should be controlled to protect the system's productivity and the
consistency of a guaranteed product quality. Virtual Engineering (VE) multi-disciplinary simulation
provides insights to explore design options and evaluate different scenarios whilst considering multiple
criteria. This paper presents how VE can be used in the design of AMS to meet the objectives for
productivity, re-configurability and process quality. Commercial case studies relevant to various industries
are presented, including personal care formulation, pharmaceuticals, material handling and packing. The
main contribution is the demonstration of how system design and VE work together to effectively support
the design challenges. The steps and key considerations are presented.
1 INTRODUCTION & PROBLEM DESCRIPTION
VE is increasingly used in the design and development of AMS. The typical approach starts with defining
the stakeholder requirements, moving onto evaluating the extent to which the generated concepts and / or
initial designs meet the requirements. This can be an intricate and time-consuming task, due to the multi-
disciplinary nature of these evaluations. It is often considered to be crucial to ensure full alignment with the
requirements to avoid costly engineering changes later in the project. VE analysis provides insights into
performance metrics and allows the evaluation of a systems behaviour for different scenarios. The latter is
becoming more important as AMS are expected to rapidly reconfigure to address changes in customer-
needs, such as in the Fast Moving Consumer Goods sector. Also, efficient evaluation can contribute to
identifying innovative concepts and de-risking less-conventional designs. The following challenges with
VE at the initial stages of system design have, however, been identified: (1) how to deal with unknowns
and uncertainty; and (2) how to avoid outdated models, version control of models and data. Initial concepts
and designs change frequently and significantly in the early conceptualisation stages, making it a challenge
to keep models up to date with the latest concepts and designs, which in turn results in additional time and
effort obligations
2 PROPOSED VIRTUAL ENGINEERING METHODOLOGY
A five-step VE methodology to handle the two aforementioned challenges is proposed:
STEP 1: Approximate the analyses objective and scope, assessing which system design aspects
require VE support and where most impact can be delivered. During this, reviewing lessons learnt
from previous projects is highly recommended.
STEP 2: Identify and map models to the analyses objectives and scope. For each model, it is
necessary to decide how to develop these in order to analyse different candidate solutions along the
timeline for the design changes and updates to support decision making. Three different approaches
Glorieux and Purse
can be used for this: (1) develop a model iteratively, continuously reconfigure and update the model
as the concepts and designs are refined; (2) develop the model as sprints when analysis is required;
(3) implement a batch of modifications at waypoints. The choice of approach should consider when
the model will be out of date and the effort required to update the model. It is important to note that
for all approaches, following a systematic approach for requirement and data capture will expedite
analyses delivery.
STEP 3: As analyses are performed in the initial stages of the system design, there is typically
uncertainty around the system to be modelled. Therefore, it is necessary to handle this uncertainty
in an explicit way to generate meaningful results. To do this, the following can be used: (1)
simplifications; (2) additional assumptions; (3) relative analysis. Whilst the first way is to simplify
the model to exclude the unknown / uncertain elements, including additional assumptions are useful
when considering best- or worst-case scenarios. The third option, relative analysis, gains insights
into emergent system behaviour.
STEP 4: Deliver the VE analysis using the developed models. It is suggested to store results, input
data, assumptions, and the model with the analysis to keep a complete configured record to be able
to revisit the analysis at a later point in time.
STEP 5: Maintaining and updating the model; it is important to adhere to one of the above
approaches for implementing future updates to the model, as it is ineffective (in long term) to
change and mix multiple approaches.
3 COMMERCIAL CASE STUDIES
A first commercial case study looks at an AMS for shampoo production system that includes dosing,
mixing, bottle filling, labelling and packing operations. The key requirement was rapid re-configurability
so as to allow changing between product variants as a responds to market trends. The consequence of this
was to balance the trade-offs between flexibility, cost and productivity. The VE method included discrete
event simulations (DES) and cost benefit analyses (CBA), which were continuously updated with new
functionality and datasets during design iterations. The insights into guiding equipment selection were
gained from identifying feasible time-budgets for production versus changeover, and also highlighted the
(pre)mixing processes to be the bottlenecks during both production and changeover. Through VE, the
design team was able to analyse the system’s productivity with different mixing process technologies as
well as changeover strategies, considering the respective costs for both. These VE explorations suggested
switching from inline to offline ingredient premixing to improve the productivity (>37%) to avoid delays
in the pre-mixing stage, with only small cost for increased effluent waste (<5%). Note, that being able to
identify the optimal concept (i.e. offline instead of inline) during early design phases is also highly valuable
from an engineering point of view.
A second commercial case study looks at the development of an AMS for packing pharmaceutical
tablets. The key innovation was reducing the amount of work-in-progress (WIP) that is scrapped during
changeovers. The VE method included thermal modelling (physics-based heat transfer analyses), DES and
CBA, where: the thermal modelling estimated part temperature during the different processes within the
system; DES estimated the throughput, the work-in-progress, and changeover time; CBA compared the
value of different designs against a baseline. Through initial relative analyses, the relationship between
WIP, productivity and temperature was characterised, which showed that active cooling can be avoided for
certain product types. These insights led to an innovative adaptive system design to deploy active cooling
only when / where necessary, and thereby increasing productivity (>2%) and reducing scrap rates (>15%).
Further exploration through VE showed an increased productivity can be realised (>10%) if a system
reconfigures with WIP in place, i.e. without having to empty the line, resulting in a relatively low cost
increment for more complex changeover automation.