1.258j lecture 20, workforce planning
TRANSCRIPT
![Page 1: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/1.jpg)
Outline 11. Problem description Problem description 2. General Approach 3. Strategic Level Case Study 4. Tactical Level Case Study 5. Operational Level Case Study
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
1
WORKFORCE PLANNING
![Page 2: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/2.jpg)
Existing New Hires Existing Scheduled/Operators (+) Scheduled Unscheduled Service
(+) Service Additions (( )+) (+)( )
Operator Operator Availability Demand
Attrition Absence Day Off Service Deletions (-) (-) (-)
Availability Less Demand GreaterThanThan ThanThan
Equal to
Additional Overtime Cost No Additional Additional Guarantee Cost Less Benefit Cost Less Benefit Cost Cost Additional Benefit Cost Additional Benefit CostCost
Nigel Wilson 1.258J/11.543J/ESD.226J 2 Spring 2010, Lecture 20
Basic Variables Affecting Manpower Level Decisions
![Page 3: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/3.jpg)
Level
t t
General Approach
Strategic LevelStrategic • workforce size • hiring planhiring plan • vacation allocation
Tactical Level ff b d k• extra staff by day off week
Operational Level • report times for unassigned extra staff
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
3
![Page 4: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/4.jpg)
•
Problem Description
By period of year: • absence hours • required extra work • attrition
Strategic Level: • workforce size • vacation allocation • hiring patterns
• Budget • Service Plan • Vacation Liability • Work Rules Work Rules • Policies
By garage and day of week:
• absence hours • required extra work • required extra work
Tactical Level: • extraboard allocation
by garage and day of week day of week
O ti l L lOperational Level: • report times for
available extraboard operators
• Work Rules • Policies
By garage and By garage and time of day:
• absence hours • required extra work required extra work
• Work Rules • Policies
operators
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
4
![Page 5: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/5.jpg)
costs scheduled hours
budget work rules
STRATEGICSTRATEGIC PLANNING
workforce size extra work required
ABSENCE open work OVERTIME available
OVERTIME i d t iOVERTIME workedstrain
missed trips RELIABILITY
l.o.s.
ridershipridership
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
5
![Page 6: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/6.jpg)
A.
Hiring Options
A. Quarterly HiringQuarterly Hiring Operator Surplus Operator DeficitExtraboard
Manpower Required (Available)
Winter Spring Summer Fall
(Available)
Time of Year
Timetable RRequirementsTi bl i Vacation Relief
Implications: • Unassigned cover time at start of timetable • Largge amounts of overtime at end of timetable • Poor reliability at end of timetable
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
6
![Page 7: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/7.jpg)
Total Unscheduled Pay
Workday-Off and Second Run Premium Pay with Associated
Variable Fringe Benefits
Guarantee and Associated Variable Fringe Benefits
Optimal Location
Fixed Fringe Benefits
Number of Extraboard Opperations
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
7
![Page 8: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/8.jpg)
Optimal Extraboard Size and Unscheduled Guarantee and Premium
100
A
Optimal Extraboard Workday-Off/
Second Run Premium
p(x) = 16% No Work Available Guarantee
B 0
0 10 20 30 40 50 60 720 SCHEDULED RUNS SCHEDULED RUNS
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
8
![Page 9: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/9.jpg)
1. Decision Variables
The Strategic Level Approach
1. Decision Variables • Workforce Size for Each Period • Vacation Allocation for Each Period • Optimal Hiring Levels for Each Period
22. Obj ti Mi i i W kf C tObjective: Minimize Workforce Cost • Scheduled Runs • E t bExtraboardd • Overtime
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
9
![Page 10: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/10.jpg)
3. Constraints
-
The Strategic Level Approach
3. Constraints • Vacation Liability
• Overtime
• Service Reliability
• PartPart-time Operation Constraintstime Operation Constraints
• Other Policy Constraints
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
10
![Page 11: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/11.jpg)
Overtime and Feedback
1. Reggular Overtime • the result of more required work than available extraboard on a
given day
2. Excess Overtime • the result of inherent inefficiency in assigning daily report times
DAILY OPERATIONS PROFILE
Open Runs/ Open Runs/ Available Coverage
Ti f D Time of Day Open Run Profile Available Cover Profile Unassigned Cover Unassigned Cover Overtime or Missed Trips
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
11
![Page 12: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/12.jpg)
The Excess Overtime Curve
• Excess overtime is a maximum when the number of required work hours exactly matches the number ofrequired work hours exactly matches the number of extraboard hours available
• Excess overtime decreases with fewer required work hours or available workforce hours
Daily Excess Overtime Curve
Excess O ti
k
Overtime Hours
-b +b0 Available Cover - Required Cover
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
12
![Page 13: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/13.jpg)
Expected Overtime for the Period
• Takes into account of variation of both the reqquired work hours and day-to-day variability of the size of the extraboard
REGULAR AND EXCESS OVERTIME
TOTALTOTAL
EXCESS
REGULAR
|
0
EXCESS
WORKFORCE - AVERAGE REQUIRED WORK
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
13
![Page 14: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/14.jpg)
Overtime Effects on Total Workforce Costs
TOTAL WORKFORCE COSTS
TOTAL COST
TOTAL OT SCHEDULED WORKFORCE
EXCESS OT
REGULAR OT
WORKFORCE
|
0
OT
WORKFORCE - AVERAGE REQUIRED WORK WORKFORCE AVERAGE REQUIRED WORK
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
14
![Page 15: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/15.jpg)
t
A Reliability Model
Open Work
Overtime Available
O A il bl
Missed Trips Overtime Worked
Yes
Operator Available Yes
No No Willingness to Work Overtime
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
15
![Page 16: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/16.jpg)
Missed Service Hours
250
100
5050
00 100 200 300 400 5000 100 200 300 400 500
Open Work (Hours)
Missed Service Hours = 0.28 x Open Work Hours
Mis
sedd
Trip
s (H
oours
) 200
150
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
16
![Page 17: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/17.jpg)
on Massachusetts Tr
Case Study
(Based on Massachusetts Bay Transportation Authority(Based Bay ansportation Authority Bus Operations)
Characteristics • Part-time workforce sized to 40% of the full-time
workfkforce
• Large variability in the required work hours • Mean Daily Absence and Extra Work: 1250 hours
• D il St d d D i ti f Ab d E t WDaily Standard Deviation of Absence and Extra Workk: 290 Hours
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
17
![Page 18: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/18.jpg)
MBTA Cost Analysis (1996)
Overtime Part-Timer Full-Timer
Wage Rate ($$/Hour) 29.04 19.36 19.36
Full Cost/Hour Worked 32.72 31.24 34.78
Marginal Cost if last extraboard used 75% of time extraboard used 75% of time
-- 41.65 46.37
Marginal Cost if last extraboard used 50% of time
-- 62.48 69.56
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
18
![Page 19: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/19.jpg)
urs)
Available Operator Hours
80008000 8 00 8500 9000 9 00 100009000 9500 10000 Available Operator Hours
10 0010500 1100011000
O = regular time costg o = regular time cost + regular OT costg g x = total cost
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
19
90009000
8500
8000
7500Trip
s (H
o
x x x o o o
7000
Mis
sed
T
6500
6000
![Page 20: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/20.jpg)
Results of Constant Hiring and Constant Vacation Constraints
Base Case
Constant Hiring
Constant Vacation
Constant Hiring & Vacation
FT Oper 1256.50 1256.50 1290.60 1315.90
PT Oper 653.90 653.90 665.70 684.80 Overtime (%) 1.50 1.50 0.90 0.30
OT cost* 1.45 1.45 0.88 0.30 Reg cost* 96.37 96.37 98.78 100.93
Tot cost* 97.82 97.82 99.65 101.23 Reliability (%) 99.60 99.60 99.80 99.90
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
20
![Page 21: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/21.jpg)
Results for Different Overtime Constraints
Base Case 1.5% OT
no OT Const
5% OT Const
1% OT Const
FT Oper 1266 1104 1202 1267
PT Oper 654 575 625 660
Overtime (%) 1.5 12.2 5.0 1.0
OT cost* 1.4 11.8 4.8 1.0
reg cost* 96.4 84.7 92.2 97.2
tot cost* 97.8 96.5 97.0 98.2
reliability (%) 99.6 97.0 98.8 99.8
* Costs are in millions of dollars per year Nigel Wilson 1.258J/11.543J/ESD.226J
Spring 2010, Lecture 20 21
![Page 22: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/22.jpg)
Tactical Level (Timetable/Rating Level)
Objective: minimize weighted sum of • overtime •• missed trips missed trips
Decision variables: allocate extra staff • by garage (area of depot) • by day of week
Inputs: • operator timetable requirements by day of week and garage• mean and standard deviation of absence and required extra
work by day of week and garage
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
22
![Page 23: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/23.jpg)
Tactical Level (Timetable/Rating Level)
Constraints: total available operators
K l hi Key relatiti onships: • requested overtime as a function of total available operators,
timetable requirements absence and required extra work timetable requirements, absence, and required extra work • missed service as a function of requested overtime
Method: heuristic or optimization method
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
23
![Page 24: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/24.jpg)
Application of Tactical Model to Single MBTA Garage
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
24
![Page 25: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/25.jpg)
variation in absence and extra
•
Tactical Level Findings
• Significant variation in absence and rrequiredequired extraSignificant work • by garage • by day of week
• Variably sized extraboard is appropriateVariably sized extraboard is appropriate • by garage • by day of weekby day of week
• Data required on absence and extraboard utilization b d d kby garage and day off wee
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
25
![Page 26: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/26.jpg)
minimize sum of
ato t etab e e e e ts t e o d
Operational Level (Daily Level)
Objective: minimize weighted sum ofObjective: weighted • overtime • missed tripps
Decision variables: extra staff report times in ranked order
• by garage (area or depot) • by day of week
Inputs: • operator timetable reququirements by time of dayayope by • known extra work by time of day
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
26
![Page 27: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/27.jpg)
Constraints: extraboard work rules
Operational Level (Daily Level)
Constraints: extraboard work rules
Key relationships: • likelihood of missed trip resulting if no cover operator
available, by time of day
Method: heuristic or optimization method
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
27
![Page 28: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/28.jpg)
Probability of Open Run Profile
0 065 0.065 A t lActual Smoothed0.060
____ ____
0 055 0.055
0.050
0.045
0.040
0.035
0.030
4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1
TIME OF DAY TIME OF DAY
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
0.025
28
![Page 29: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/29.jpg)
Unexpected Absences by Day-of-Week
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
29
![Page 30: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/30.jpg)
Expected Weighted Uncovered Open Work
Day of Week DHS HS FLAT Monday 53.34 54.96 53.12
Tuesday 19.36 19.18 19.42
Wednesday Wednesday 41 42 41.42 41 31 41.31 41 89 41.89
Assumes 6 FTOs, 4 PTOs available on extraboard
Key:DHS = d d h ifi b t DHS day and hour specific absence rates HS = assumes hour specific absence rates only FLAT = assumes constant absence rate for all days and hours
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
30
![Page 31: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/31.jpg)
Evaluating Current Practice: Weighted Uncovered Open Work (Hours)
Date Rep. Oper. (FTO-PTO)
Actual Rep. Model Results
6/29 11-7 36.1 26.9
7/067/06 3-03 0 118 1 118.1 112 3 112.3
7/13 6-6 64.0 54.3
7/207/20 8 128-12 40 1 40.1 22 0 22.0
7/27 10-5 53.0 36.6
Data are for 5 consecutive Mondays for a specific MBTA garage
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
31
![Page 32: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/32.jpg)
Actual vs. Recommended Report Times MONDAY, 7/13 MONDAY, 7/27
Actual Recommended Actual Recommended 4.454.45 4.304.30 4.304.30
5.00 5.00 4.30 5.30 4.45 5.45 5.00 5.00
6.00 6.00 5.00 6.00 6.00 5.00
6.00 5.30 5:30 7.00 5.45 7.00 6.00 6.00 7.00 6.00 6.00 7.00 6.00 6.00 8.00 6.00 8.00 6.00
13.45 6.15 14.00 6.30 14.00 8.00 14.15 12.00 15.30 13.00
15.45 13.45 18.15 14.00 20.00 14.00
14.15 14.45
1.258J/11.543J/ESD.226J 16.00Nigel Wilson
Spring 2010, Lecture 20
32
![Page 33: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/33.jpg)
Operational Level Findings
• Significant improvements possibleSignificant improvements possible • reduced overtime • reduced missed tripps
• Single set of ranked report times can be used across allll weekdkdays andd seasons ffor eachh garage • separate ranked report times required for Saturdays, Sundays
• Constant absence rates can be assumed • byy hour of dayy • by day of week
Nigel Wilson 1.258J/11.543J/ESD.226J Spring 2010, Lecture 20
33
![Page 34: 1.258J Lecture 20, Workforce planning](https://reader034.vdocuments.us/reader034/viewer/2022052016/62864ea0eaeecb69ca22bf0c/html5/thumbnails/34.jpg)
MIT OpenCourseWarehttp://ocw.mit.edu
1.258J / 11.541J / ESD.226J Public Transportation SystemsSpring 2010
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.