Research Overview

1
Research Overview

• Christian Hicks

2
Main Research Areas

• Simulation / modelling of systems
• Scheduling / planning and control
• Manufacturing Layout
• Supply chain management
• IT implementation

3
Other areas

• Company classification
• Manufacturing strategy
• Business Process Analysis
• Benchmarking in the semiconductor industry
• Web-based teaching

4
Capital Goods Companies Generic Issues

• Products are highly customised and are produced
  on a make, or engineer to order basis.
• Lead time reduction and cost increasingly
  important.
• International competition effective and
  efficient use of resources is very important.
• Product offering broadened over recent years to
  include service elements.
• Complex and dynamic supply chains.
• Production facilities include jobbing, batch,
  flow and assembly systems as well as
  construction.

5
(No Transcript)
6
(No Transcript)
7
Simulation of Systems
8
(No Transcript)
9
Key Features

• Large scale model allows whole manufacturing
  facilities to be represented.
• Models facilities, products, processes, layout
  and planning and control systems.
• Many product families can be represented with
  shallow, medium or deep product structure.
• Hierarchical description of products and
  resources.
• Allows variety of planning and control methods to
  meet local requirements.
• Integrated with scheduling and layout
  optimisation tools.
• Comprehensive stochastic modelling.

10
Schedule Optimisation
11
Schedule optimisation using Genetic Algorithms
12
Initial Schedule
13
(No Transcript)
14
New schedule from GA
15
Stochastic Planning Methods

• Risk management is an important issue in capital
  goods companies. Developed stochastic models of
  risk due to uncertainties in process times.
• Developed methods that either meet a service
  target or minimise the combination of earliness
  and tardiness costs.
• Investigated approaches for infinite capacity,
  finite capacity and dynamic scheduling cases.
• Planning methods investigated and validated
  through simulation modelling.

16
Manufacturing Layout

• Clustering
• Matrix-based methods
• Similarity coefficient methods
• Optimisation
• Genetic Algorithm
• Simulated Annealing

17
Genetic Algorithm Procedure
18
Placement Algorithm
19
Initial Layout
20
Total rectilinear distance travelled vs.
generation (brown field)
21
Resultant Brown-field layout
22
Total rectilinear distance travelled vs.
generation (green field)
23
Supply Chain Management
24
Supply Chain Management

• Modelled business processes using SSADM
• Company structures range from vertically
  integrated to project integrators that outsource
  all manufacturing.
• Important factors include available capital,
  risk, potential utilisation of plant,
  capabilities, flexibility.
• Three stages of interaction with customers
  marketing, tendering and contract execution
• Normal / radical design
• Functional vs. technical specifications
• Procurement decisions made by customers,
  designers, procurement departments

25
Summary of contributions

• Planning, control and layout problems in capital
  goods companies.
• Outcome first large-scale simulation model of
  manufacturing in capital goods companies
• Scheduling complex products in deterministic and
  stochastic environments.
• Developed first optimisation techniques.
• Layout analysis and optimisation.
• Developed integrated tool.
• Supply chain management in capital goods
  companies.
• New models proposed and linked to strategic
  issues.

26
Classification
27
(No Transcript)
28
Simulation
29
(No Transcript)
30
(No Transcript)
31
Representation of Resources
32
Representation of Products
33
Stochastic Simulation

• Several random number generators Knuth, Wichman
  Hill, SunOs.
• Normal polar form of Box-Muller (Marsaglia and
  Bray 1964) Beta (Press, et al. 1989, p188),
  Gamma (Press, et al. 1989, p228), Poisson (Press,
  et al. 1989, p230) as well as Log normal,
  Multi-modal, Exponential, and empirical (based
  on historical data).
• Full / fractional factorial designs
• ANOVA / Regression analysis

34
Scheduling
35
Chromosome representation
36
Crossover Operations
37
Mutation Operations
38
Fitness function
Minimise ? Pe(EcEp) ? Pt(Tp) Where Ec
max (0, Dc - Fc) Ep max (0, Dp -
Fp) Tp max (0, Fp - Dp)
39
An Example of Production Plan
40
Layout
41
Rank Order Clustering (King 1980)
42
Case Study

• 52 Machine tools
• 3408 complex components
• 734 part types
• Complex product structures
• Total distance travelled
• Direct distance 232Km
• Rectilinear distance 642Km

43
Total rectilinear distance travelled vs.
generation (green field)
44
Resultant green field layout
Note that brown field constraints, such as
walls have been ignored.
45
Supply Chain Management
46
Collaborating Companies

• NEI Parsons
• AMEC Offshore
• NEI International Combustion
• Clarke Chapman
• Wellman Booth
• Control Systems
• Reyrolle (VA Tech)

47
Supply Chain Management

• Identified the characteristics of the companies
  in terms of products, processes, markets, level
  of outsourcing etc.
• Investigated buyer/supplier relationships in
  terms of supplier base, strategic alliances,
  partnership and single sourcing agreements etc.
• 3 stages marketing, tendering, contract
  execution
• Physical / non-physical processes,
• Differing levels of vertical integration
• Procurement often reactive rather than strategic

48
SCM (continued)

• Majority of controllable cost committed at the
  design stage.
• Normal / Radical design
• Established / ad-hoc business processes
• Product offering broadening shift from just
  hardware to retrofit, service and operations.
• There are high levels of uncertainty and sparse
  knowledge, particularly at the tendering stage.
• Tendering is often subject to severe time
  pressure and resource constraints.

49
(No Transcript)
50
(No Transcript)
51
(No Transcript)
52
IT Implementation in Small Companies

• Selection and implementation of Accounting,
  planning and control systems and EDI in surgical
  shoe manufacturing company.
• Selection and implementation of a cost estimating
  system in a precision machining company.
• Selection and implementation of accounting and
  order processing systems in an electrical cable
  manufacturing company.

53
Benchmarking in the semiconductor industry

• Benchmarking survey of four Siemens plants
  WhiteOak (USA), ProMos (Taiwan), Siemec (Germany)
  and NTS (England).
• The design, construction, and operation of the
  plants are different leading to different
  consequences in terms of capital and operating
  costs, quality, etc.
• The research aims to establish the relationship
  between capital cost, construction lead-time and
  operating costs and performance.

54
Postgraduate Students

• Dong Ping Song Stochastic Models in Planning
  Complex Engineer-to-Order Products
• Pupong Pongcharoen Genetic Algorithms for
  Scheduling in the Capital Goods Industry
• Tony Wells Benchmarking in the Semiconductor
  Industry
• Fouzi Hossen Risk Management in Capital Goods
  Companies
• Thanawat Muangman Supply Chain Management in the
  Capital Goods Industry

55
Research Grants

• An investigation into the design change
  management process and the relationship to
  resultant cost of change for capital equipment at
  BNFL, 160,000, 2003, (CH/PMB/WO)
• Harris Sheldon TCS, 2000-2, 81,408,
  discontinued, (RID/CH/PMB)
• Haani Cables TCS, 98-2000, 141,760,
  (CH/TMcG/MG/PMB)
• SCM at Wellman Booth, 1997, 2,000, (CH/TMcG)
• Mectonics Instruments TCS, 95-7, 65,000
  (CH,CFE,WH)
• Innovation in Design, 1996-9, 10,000 (JNF/CH)
• JC Peacocks Sons TCS, 91-5 122,552 (GRJ/CH)
• Total income 490,000 Principle investigator
  368,000