hrm and sd

20
Supporting Long-term Workforce Planning with a Dynamic Aging Chain Model: A Case Study from the Service Industry with two more recent additions Andreas Größler

Upload: andreas-groessler

Post on 03-Jul-2015

418 views

Category:

Documents


0 download

DESCRIPTION

A presentation that talks about using system dynamics in an HR management context

TRANSCRIPT

Page 1: HRM and SD

Supporting Long-term Workforce Planning

with a Dynamic Aging Chain Model:

A Case Study from the Service Industry

with two more recent additions

Andreas Größler

Page 2: HRM and SD

The case company: air traffic control

© spiegel-online.de

Page 3: HRM and SD

Long-term personnel capacity planning

as a crucial success factor

Quote Eurocontrol:

• “Manpower Planning (MP) is

necessary to meet the strategic

objective:

‘the provision of the right number of staff, with the right

qualification, at the right time and in the right place to meet

business requirements’”.

Page 4: HRM and SD

Goals of modelling and simulation

project

• Conduct a structural analysis of the existing long-

term personnel planning process for air traffic

controllers;

• provide a dynamic analysis of the existing planning

policies;

• construct a scenario-tool to improve the existing

planning policies as well as the established risk

management approach accompanying the existing

processes.

Page 5: HRM and SD

A simple system dynamics model of

capacity planning

#ATCOs at

training

#ATCOs

operative recruiting graduating leaving

recruitment time training time productive time

AIR TRAFFIC requirements

#ATCOs

needed

Page 6: HRM and SD

Some results from simple model

#ATCOs operative

6,000

4,500

3,000

1,500

0

2006 2010 2014 2018 2022 2026 2030 2034

Time (Year)

AT

CO

#ATCOs needed

600

450

300

150

0

2006 2010 2014 2018 2022 2026 2030 2034

Time (Year)

AT

CO

Independent from the scenario used, there are variations in some key

variables over time that are not easy to understand.

AIR TRAFFIC requirements)

6,000

4,500

3,000

1,500

0

2006 2010 2014 2018 2022 2026 2030 2034

Time (Year)

AT

CO

No Cycle With Cycle

Page 7: HRM and SD

Time lag through training process

Start of

process

Signing of

contract

Start of

training

Varies for

each trainee

(Ø 24 Month)

Identical for

all trainees

(15 Month) Varies for

each trainee

(Ø 6 Month)

End of

training, start

of OJT

End of OJT

Varies for

each trainee

(Ø 6 Month)

t = 0 t = 6 t = 12 t = 27 t = 51

Page 8: HRM and SD

Resulting delay behaviour: average is

longer than they think

Ordered ATCO

Signed ATCO

ATCO in IT

ATCO in OJT

ATCO

OJT 18 Months

0 10 20 30 40 50 60 70 80 90 100

Time (Month)

100

75

50

25

0

pe

rso

n

100

75

50

25

0

0 10 20 30 40 50 60 70 80 90 100

Time (Month)

pe

rson

OJT 24 Months

t≈68 t≈58

Page 9: HRM and SD

Results from client‘s perspective

• A more detailed planning paradigm can be

implemented (group level instead of centre level);

• the personnel planning cycle can be repeated several

times a year instead of only going through the process

once a year;

• the risk management can be complemented by some

quantitative scenarios that are provided almost in real-

time;

• intensified communication between all stakeholders;

• the new scenario tool can act as a learning platform for

the case company as it integrates the experience and

perspective of several departments.

Page 10: HRM and SD

Addition 1: Chains of entities…

Page 11: HRM and SD

A general issue resulting from the case

There is some confusion about the different types of supply

lines (quotes from HRM review process):

“The firm is a logistics service provider and is a service firm. The

issues at this firm are similar to issues faced by a service firm.

Service firms, similar to a logistics provider do not have a physical

product and definition of inventory is very different.”

“Supply chain in service firms are different and have their own

specific issues. References relate to service chain issues in a

service firm…”

Page 12: HRM and SD

Structural similarity of the three types

of chains

Material Work in

progress

Finished

goods purchasing fabricating assembling shipping

Proposal Draft Final

report accepting drafting finalizing delivering

Newly

hired

In

training

Fully

productive hiring starting

training

finalizing

training leaving

Physical goods supply line

Service supply line

Personnel supply line

Be aware of the ethical issue qualitative vs. quantitative individualism (Simmel)

Page 13: HRM and SD

Prototypical behaviour of chains

20

15

10

5

0 3 3 3 3 2

2

2 2

2

2 2 3 3 3 3 2 2 2 2 2 2 2 3

3

3

3 3

3

3 3 1 1

1

1 1

1

1 1 1 1 1 1 1 1 1 0 5 10 15 20 25 30 35 40 45 50

time steps

en

titie

s

1st stock 1 1 1 1 1 1 1 1 1 1

2nd stock 2 2 2 2 2 2 2 2 2

3rd stock 3 3 3 3 3 3 3 3 3

Page 14: HRM and SD

Perceived differences and structural

similarity

• Perceived differences of the three types of chains, in

particular regarding

– Utilization of “production” capacities

– Premature outflow from the chain

– Divisibility of entities

• Because of structural similarity, differences are mainly

caused by

– Inappropriate mix of supply line elements with attributes of these

elements (“co-flow”)

– The three types of supply lines regularly are located at different

organisational levels

• “Strategic architecture” (Warren, 2007)

Page 15: HRM and SD

Addition 2: Forms of delays…

Page 16: HRM and SD

Female professors task

• The analytics of a gender quota

• Dutch university: balance number of male and female

professors

• Participants have shown gross mis-estimations

(Bleijenbergh et al. 2011)

• Influence of political loadedness of task?

Page 17: HRM and SD

Discrete vs. continuous delays

= two experimental groups

x x

to t* t to t*

t

input system response

Page 18: HRM and SD

Task structure in system dynamics

notation

Male

professors

Female

professors

hiring male profs

hiring female profs

leaving male profs

leaving female profs

avg time at

university percentage female profs

hirings necessary

Ini Male

Ini Female

Page 19: HRM and SD

Estimations do not differ between

experimental groups

32%

41%

27% outside bounds (% < 50 or > 100)

wrong estimate (diff. > 5 years)

correct estimate (diff. <= 5 years)

No statistical differences between experimental groups for values

of estimations participants do not differentiate between discrete

and continuous delays.

Page 20: HRM and SD

Average error between experimental

groups differs a lot, though

0

1

2

3

4

5

6

7

8

9

discrete continuous

years

Significant statistical differences between groups for goodness of estimations (estimations compared to “true” solutions derived from respective simulation model – discrete

vs. continuous delay version).