Lean Six Sigma Methodology in Thames Valley Probation – The

Journey

Elizabeth Medhurst, Business Process Improvement Manager, Thames Valley Probation

ORS – CJSIG 14 November 2011

Where we are

• Berkshire, Buckinghamshire, Oxfordshire.

• Combined population of about 2.1 million people.

Who we are

• About 700 staff• More than 60% work directly with offenders. • Seven Prisons• Five Approved Premises (hostels)• Twelve Magistrates' Courts

– & three Crown Courts

• Nine main offices across three counties• Plus supporting roles - training, information technology,

finance, HR and Comms etc

• We are the major providers of services to Magistrates and Crown Courts - providing 8,000 pre-sentence reports per year

• We manage around 8,500 offenders at any one time. Almost 300,000 hours of unpaid work are carried out by offenders each year - that’s about £1.5 million worth of free work

• Protecting people from the risks posed by a relatively small number of highly dangerous offenders is a top priority

• The ability of offenders to change and rehabilitate with the appropriate support and guidance is central to the probation ethos.

In the beginning ...

• Business Process Improvement programme started Sept 2008, initiated by Corporate Directors

• Value for Money programme• Moving from fire-fighting to self-determination• Three BPI managers trained as Master Lean Belts, all

with operational management experience.

Initial Approach• Early priorities identified• Three separate DMAIC projects started, emphasis on

identifying and removing waste• Court work, Pre-sentence Reports in custody and

allocation to groupwork – over-processing and re-work were the most common wastes

• Significant Challenge – how to communicate and engage wider organisation.

A communication problem?

From http://www.leanblog.org/2011/06/a-collection-of-dilbert-cartoons-on-lean-andor-six-sigma/

Creating sustainability

• First projects had started to get below the surface of what was really happening

• Invaluable to have “asked the parts” from the beginning (Deming)

• Work plan produced for the next round of improvements

• BPI managers reduced from three to two• Initially intended to complete projects and then

publish the results in process maps.

Creating sustainability• With greater clarity, change of approach needed - the jigsaw• Decision to map everything “as is” then review work plan• Intelligent Operations Manual (How2) created in Nimbus

Control• The problem areas became clearer as a result• Having a platform to build on, used by all staff helps to create

sustainability• The How2 manual also created improvements by clarity and

standardisation.

How2 – the intelligent operations manual – information within 3 mouse clicks

Detailed diagrams for complex areas which require standardisation

Getting serious ...• Cuts in budget along with unit costing exercise led to

initiating a Lean Offender Management project• Objectives to maximise meaningful offender contact

and reduce the cost of these activities by at least 20%

• Facilitated workshops across all areas attended by a cross-grade staff group ensured valuable input by process experts.

Implementation

• Thorough data analysis after the event validated the areas where the most savings would come from

• Put into the business plan for 2011/2013• Some areas handed to process owners to deal with• Large areas subject to formal DMAIC projects.

PSR project

• Will be black belt accredited on completion• The need to reduce the cycle times of reports and

adapt to changes in court listings and performance requirements were key drivers

• Chose high volume, variable office, on the basis of being able to transfer successfully to wider organisation

• Currently in Improve cycle.

Force Field Analysis – PSR project

17

Critical to Quality Factors

Pre-Sentence Reports are divided into three categories - Oral, Fast Delivery (FDR) and Standard Delivery (SDR). Oral reports are produced and delivered verbally on the day of requests, and FDRs and SDRs after an adjournment.

In order to meet the key customer requirement to increase the amount of Orals, and reduce the overall time for report completion for adjourned reports, the following factors have been identified as critical to quality:

• Speed - The time in working days from the report being requested by the court to the interview taking place.

• People - The ability and confidence of the court duty officer to identify the correct type of report and advise the court at the point of request – compliance to process.

• Capacity - The availability of report writers to meet overall demand and also to be available for interview as early as possible.

• Accuracy – ensuring that the right level of report is identified at the first opportunity, with supporting information to pass to the report author.

18

Measure/Analyse – Capability Assessment of FDRs interviewed within 5 days of request showing Sigma rating of 1.2

Activity Characteristic Standard Frequency

Fast Delivery Reports Working days to interview First 24 sentenced during Jan 11

UpperDistribution of X class

34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 limit f

Xs

la

u X 14 1

di X 12 1

vi X X X X 10 4

dn X X X X 8 4

I

X X X X X X 6 6

X X X X X X X X 4 8

Remarks

Rm Based on all observations

-

e Sigma = + 1.5 = + 1.5 =

gn

a Yield =

Rg

nI

vo

M

n = 2

UCLx = X-bar + 3 sigma hat =

LCLx = X-bar - 3 sigma hat =

Constantsn D3 D4 d2

2 N / A 3.268 1.1283 N / A 2.574 1.6934 N / A 2.282 2.0595 N / A 2.114 2.326

3 1013 4 4 09 6 9 119 8 2 74 6 6 6X 5 7 5 4 3 8

Process Improvement Chart XmR Process owner: John Ennis, Senior Probation Officer Chart owner:

6.2 2.6X-bar = sigma hat = R-bar / d2 =2.9R-bar =

mR 2

1.2

39.19%

5.5 6.2083

-----------------------

2.58

7

LCLR = D3 R-bar = N / A

UCLR = D4 R-bar = 9.514.0

-1.5

9 0 4 33 3 2 22 1 1 5 4 2 0 0 3 1 6 5 2

Total 24

24

standard - X-bar

----------------------------------------

sigma hat

Nominal

Sheet number: 01Apr11CAPJohn Ennis

Upper tolerance 5.5

Lower tolerance

0

2

4

6

8

10

12

14

16

18

20

0

1

2

3

4

5

6

7

8

9

10

ANOVA for court duty officers when exploring compliance to process12

3

4

5

6

7

89

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869

B C D E F G H

C DO1 C DO2 C DO3 C DO4 Mean9 7 0 02 9 2 56 10 6 58 3 0 09 0 0 0

13 8 5 6Mean 7.83 6.17 2.17 2.67 4.71

Total SoS = 356.96 with 23 df

Remove Error5.67 8.67 2.67 3.335.67 8.67 2.67 3.335.67 8.67 2.67 3.33

10.00 3.67 1.67 2.0010.00 3.67 1.67 2.0010.00 3.67 1.67 2.00

Remove Error SoS = 204.96 with 7 df

Officer7.83 6.17 2.17 2.677.83 6.17 2.17 2.677.83 6.17 2.17 2.677.83 6.17 2.17 2.677.83 6.17 2.17 2.677.83 6.17 2.17 2.67

Officer SoS = 135.13 with 3 df

Court5.08 5.08 5.08 5.085.08 5.08 5.08 5.085.08 5.08 5.08 5.084.33 4.33 4.33 4.334.33 4.33 4.33 4.334.33 4.33 4.33 4.33

Court SoS = 3.38 with 1 df

Unexplained8.04 3.04 2.04 1.381.04 5.04 4.04 6.385.04 6.04 8.04 6.382.71 4.04 3.04 2.713.71 1.04 3.04 2.717.71 9.04 8.04 8.71

Unexplained SoS = 152.00 with 16 df

Interaction2.17 6.83 4.83 5.002.17 6.83 4.83 5.002.17 6.83 4.83 5.007.25 2.58 4.58 4.427.25 2.58 4.58 4.427.25 2.58 4.58 4.42

Interaction SoS = 66.46 with 3 df

ANOVA tableSource SoS df MS F Ratio P value 1 - POfficer 135.13 3 45.04 4.74 0.01497 98.50%Court 3.38 1 3.38 0.36 0.55949 44.05%

Interaction 66.46 3 22.15 2.33 0.11284 88.72%Unexplained 152.00 16 9.50

Total 356.96 23

Days to interview for 4 officers

5.08

4.33

S loug h C ourt

Maidenhead C ourt

20

Fishbone diagram of root causes

Next steps

• Development work with staff• Move resources around – optimize appointments• Create visual management for workflow as part of

control measures• Achieve sigma level of 6 for the process.

Lean Six Sigma in TVP

• A continuous improvement approach as opposed to “doing” Lean Six Sigma.

• Cashable savings identified over all completed projects approximately £850k p.a – 3.5% of budget – with more savings to come

• Non-cashable savings are significant and have freed up time for direct value tasks

• One BPI manager remaining in post• It’s never finished – the journey continues!

Thank you for listening – questions and discussion welcome.

Elizabeth.medhurst@thames-valley.probation.gsi.gov.uk