European Funding to help SMEs improve using six sigma

1
European Funding to help SMEs improve using six
sigma

• Tony Fouweather, Shirley Coleman Andrew Thomas

2
Background

• SME is defined by the European Union as an
  independent company with fewer than 250 or fewer
  employees and either an annual turnover not
  exceeding 40 million or a balance sheet not
  exceeding 27 million
• ISRU gained funding from EU to help SMEs
• ERDF -Measure 2.5 (ERDF Capital Revenue) This
  Measure provides specialised support to SMEs who
  are operating within defined clusters and sectors
  and provides intensive assistance to improve
  their competitiveness.
• ESF - This Measure complements Measure 2.5 by
  inviting organisations to run customised training
  and development packages.

3
Six sigma and SMEs

• Six sigma can improve efficiency and
  competitiveness
• SME are at a disadvantage and often lack spare
  capital or staff to implement it
• Consultants can be expensive
• Larger companies can afford to train in-house six
  sigma specialists or hire a consultant

4
ISRU six sigma training package

• Funded up to 80 of cost for SMEs
• Flexibility e.g. 1 day per week
• Six sigma project from within the company
• Leads to implementation on a wider scale
• Based on DMAIC strategy developed in 1980s

5
DMAIC - Six sigma strategy

• Define - Defining the problem, the project goals,
  the project scope and the overall strategy.
• Measurement Deciding what quality
  characteristics measure current process
  performance and to measure progress towards the
  project goals. Verifying the accuracy, precision,
  repeatability and reproducibility of the
  measurement systems. Determining the current
  process / system capability.
• Analysis - Identifying and quantifying sources of
  variation.
• Improve - Removing causes of variation,
  discovering variable relationships and
  establishing optimum operating levels.
  Development of an implementation strategy.
• Control - Implementing controls and determining
  improved process/system capability. Holding the
  gains by ensuring that all the changes are fully
  documented and become part of the standard
  operating procedures.

6
Case study 1 FM foods

• Small local Bakery based in Sunderland with 30
  employees
• The delegate view of the training package The
  Six Sigma training gave us a set of tools which
  allowed us to improve the efficiency of our
  packing line for one of our most difficult
  products

7
Definition of the problem

• Product flow through the Ishida machine varies
  considerably when packing pineapple and the
  reject rate is high. The aim of the six-sigma
  project is to improve the operating efficiency of
  the Ishida
• Ishida is an integrated weighing and sealing
  machine designed for packaging a wide variety of
  high volume non fluid materials

8
Cause and Effect Diagram
9
Measurement Phase

• Data collection sheet designed to collect as much
  information relating to as many of the variables
  as possible
• Gave good overall picture of the process
  variables
• Response variable chosen is number of bags packed
  in half hour period
• Average of 137.1 bags per half hour was packed at
  beginning of project

10
Analysis Phase
11
Analysis Phase

• Operating procedures varied
• Stickiness of pineapple can vary depending on how
  its stored and for how long its stored
• One operator commented that when she starts to
  run the machine she takes pineapple from the
  bottom 2 boxes of the stack
• Another mixed the drier and sticky pineapple
  together
• Another suggested shrink wrapping the boxes
  before putting the lid on

12
Analysis Phase

• Regression analysis carried out on the collected
  data to identify significant variables
• Looking for variables with significant impact on
  the response variable
• Designed Experiment set up concentrating on
  suspected significant factors
• Non-controllable factors also assessed using
  regression analysis

13
Improve Phase

• Ishida settings were thought to be important
• Initially the company were taking all 4 settings
  together as one factor (in the form of a 4 digit
  number)
• Split up For example, 5993 would be recoded as
  A5, B9, C9, D3
• Settings currently used by the operators were
  incorporated into the experiment

14
Minitab output (Regression)

• Significant factors/interactions are indicated by
  smaller p values

Estimated Effects and Coefficients for Amount
(coded units) Term Effect Coef SE Coef
T P Constant 51.250 3.910
13.11 0.000 A -2.250 -1.125 3.910
-0.29 0.781 B 19.750 9.875 3.910
2.53 0.036 C -11.250 -5.625 3.910
-1.44 0.188 D 15.750 7.875 3.910
2.01 0.079 AB or CD 9.500 4.750 3.910
1.21 0.259 AC or BD 11.500 5.750 3.910
1.47 0.180 AD or BC -18.500 -9.250 3.910
-2.37 0.046 S 15.6405 R-Sq 73.17
R-Sq(adj) 49.70
15
Ishida setting main effects plots

• Significant main effects indicated by large
  change between the levels

16
Interaction between Core vibration (factor A) and
Radial Vibration time (Factor D)

• Significant interaction as lines not parallel
• Best setting is factor A set at 5 with D set at 3
• Confounded with interaction BC so effect may be
  due to either of these or combination of both

17
Interaction between Core Vibration Time (factor
B) and Radial vibration (factor C)

• Significant interaction as lines not parallel
• Best setting is factor B set at 9 with C set at 5
• Confounded with interaction AD so effect may be
  due to either of these or combination of both

18
Control Phase

• Core vibration (factor A) 5
• Core Vibration Time (factor B) 9
• Radial vibration (factor C) 5
• Radial Vibration time (Factor D) 3
• Dramatic improvement in the process was noted

19
Benefits

• 6871 direct saving
• Decreased reject rate
• Improved product flow
• Improved awareness of quality
• Establishing the causes of process variation
• Increase in confidence and ability
• Better team working

20
Case study 2 Thomas Swann

• Thomas Swan Co. Ltd based in Consett, County
  Durham
• Chemical manufacturers who manufacture and sell
  speciality chemicals for the performance and fine
  chemicals markets
• Training provided modelling techniques to improve
  a process

21
Define Phase

• Problem with drying process
• Need to predict the time taken for a batch of the
  chemical to dry
• 0.2 water is the specification set by the
  customer
• Factors - original start weight, drying
  temperature, percentage water present at start of
  process
• Boxing system was also improved during define
  phase

22
Measurement Phase

• Sample of collected data

23
Scatterplot
24
Analysis Phase

• The initial model
• Minutes to dry 431 - 59.9 WATER 0.0776
  weight - 3.85 temp
• The equation refers to elapsed drying time,
  rather than the time taken to dry. Hence, the
  elapsed drying time is least when the of water
  is most. The equation is purely mechanical
  assuming all points are independent.

25
Regression Analysis

• Significant factors with a low p-value (less than
  0.05)
• R square (adjusted) value of 97.7 indicates most
  of variation is explained by the variables in the
  model, so no need to consider interactions

Regression Analysis MIN versus WATER, weight,
temp Predictor Coef SE Coef T
P Constant 430.92 66.26 6.50 0.000
WATER -59.920 1.506 -39.80 0.000 weight
0.07762 0.03060 2.54 0.016 temp
-3.846 1.027 -3.74 0.001 S 11.9367
R-Sq 97.9 R-Sq(adj) 97.7
26
Improve Phase Control Phase

• Drying time (minutes) 114 0.0707 wet wt
• Occasionally the percentage water was just above
  the 0.2 specified
• Drying time (minutes) 10 minutes 124 0.0707
  wet wt
• Drying time can now be estimated accurately
  leading to shorter cycle times

27
Benefits

• An extra batch each week can be produced in the
  saved time
• 6,000 profit per week
• 300,000 per year

28
Conclusion

• Simple statistical techniques allow improvements
  in efficiency and competitiveness
• Gain more control over the processes
• Substantial increases in output / profit
• Better skilled staff
• Embed advanced statistical methods
• Improved team-working skills

29
Project Deliverables

• The project is coming to an end and ISRU have
  helped 23 local SMEs for at least 25 days each
• Need to help 2 more by end of year
• In addition basic advice to 120 companies
• 25 SMEs to implement outcomes of assistance
• 25 SMEs to improve environmental performance
• 25 SMEs to enhance applications of ICT
• 15 SMEs to introduce new or improved product
• 15 SMEs to implement process improvements

30
Applying training model at MEC

• Cardiff MEC has access to various sources of
  funding
• Can set up a similar scheme to help their local
  SMEs implement six sigma strategy
• Improve efficiency and competitiveness