Dynamic Pedestrian and Vehicular Modelling

1
Dynamic Pedestrian and Vehicular Modelling

• J. MacGregor Smith M. Blakey Smith
• Department of Mechanical and Industrial
  Engineering Facilities Planning
• University of Massachusetts
• Amherst MA 01003 http//www.ecs.umass.edu/mie/fac
  ulty/smith/

2
Overview

• Methodology
• Representation
• Analysis
• Synthesis
• Case Studies
• Newton-Wellesley Hospital Campus
• Engineered Polymers Warehouse Facility
• Automated Teller and Walkup Facility

3
Basic Methodology

• Representation
• Step 1.0 Define Customer Classes
• Step 2.0 Define Routing vectors
• Step 3.0 Define Distance and Flow Matrices
• Analysis
• Step 4.0 Topological Diagrams
• Step 5.0 Layout Alternatives
• Step 6.0 Flow Analysis
• Synthesis
• Step 7.0 Evaluation of Alternatives
• Step 8.0 Synthesis
• Step 9.0 Implementation

4
Representation

• Step 1.0 What customers (patients, staff,
  visitors) are moving through the facility?
• Step 2.0 Define the route sheets for the
  customer classes
• Step 3.0 From-To Charts, Distance and Flow
  Matrices P(i,j) D(i,j)

5
Analysis

• Step 4.0 Generate the topological diagram
  relating the route sheets and the physical
  facility.
• Step 5.0 Generate the alternative layouts
• STEP/MAFLAD
• Step 6.0 Generate the analytical and simulation
  models for analysis
• QNET/ARENA

6
Synthesis

• Step 7.0 Evaluation of Alternatives
• Step 8.0 Synthesis of Results (sensitivity
  analysis)
• Step 9.0 Implementation of Plans

7
Design Issues

• What are the fundamental design and performance
  variablesinvolved in designing a circulation
  system?
• How are these fundamental design (d), performance
  (p), and contextual (c) variables related?
  Pf(c,d)

8
What is Congestion?

• Congestion occurs mainly as a result of increased
  number of pedestrians and vehicles competing for
  the limited space of a corridor or roadway
  segment.

9
Empirical Model

• The service rate (speed) decays within a corridor
  
• Since there is a finite amount of available space
  within each corridor, the density of pedestrians
  reaches an upper limit (jam density).

10
Performance Measures

• Flow (q) Output volume or throughput
• Density (k)the number of customers or vehicles
  travelling over a unit length
• Pedestrian Speed (?)
• Time (t)

11
Design Variables

• Length (L) of the corridor or highway segment
• Width (W) of corridor or highway segment.
• Capacity (C) 5 LW
• Other variables of interest e.g.
• pavement materials, patterns, etc.
• grades, stairs, etc.
• geometric curves, etc.

12
Contextual Variables

• Input Volume (?) total number of pedestrians
  that enter a circulation segment during a given
  time interval
• Other contextual variables
• obstacles,
• weather conditions
• climate
• wind conditions

13
Building Blocks Methodology

• Public Buildings
• Airports
• Courthouses
• Hospitals
• Malls
• Campuses
• Freeways

14
Linear Model

• A average pedestrian or vehicle speed
• C capacity of the corridor or highway segment C
  f(L,N)

15
Exponential Model

• ? scale parameter
• ? shape parameter

16
Empirical Curves of Pedestrian Stairwell Flows
(after Fruin)
17
General Models of Pedestrian Flows
18
Representation of Facilities

• Floor Plan/Section Graph Representation

19
Newton Wellesley Hospital Campus
20
Typical Hospital Floor
21
Routes of Pedestrian Travel
22
Routes of Pedestrian Travel
23
3d representation of the Campus
24
Pedestrian Route Analysis
25
Synthesis
26
Optimal Routes of Pedestrian Travel
27
Optimization of the network
28
Engineered Polymers Inc.

• Warehouse capacity analysis
• Dynamic material handling design
• Layout and equipment needs
• Forecast space utilization over time
• Bottleneck analysis
• Over to Simulation model animations

29
Volume of Boxes
30
Warehouse Capacities

• Raw materials Semi-Finished

31
Total Warehouse Capacity
32
Average Turnaround Time
33
Equipment Utilization
34
Extruder Utilization
35
Holyoke Power and Light

• Pedestrian vehicular layout alternatives
• Horseshoe Counter with 1 ATM
• Horseshoe Counter with 2 ATMs
• Linear Counter
• L-Shaped Counter
• Animations of different alternatives
• Summary of results

36
1 Horseshoe Counter Layout with one and two
drive-ups
37
2 Horseshoe Counter Layout
38
3 Linear Counter Layout with two drive-ups
39
4 L-Shaped Layout
40
Basic Data

• Cashiers
• 128 drive-ups /day, Ave Time 1 min 9 sec
• 290 walk-ins /day, Ave Time 1 min 18 sec
• Credit
• 12 walk-ins /day, Ave Time 6.42 min
• 27 call-ins / day, Ave Time 3.78 min
• Customer Accounts
• 36 walk-ins /day, Ave Time 4.30 min
• 37 call-ins /day, Ave Time 2.54 min
• Water Dept 5 of 8100 customers /month

41
Assumptions

• 70 customers enter via parking lot, 30 enter
  via Suffolk Street
• If cashiers are backed up, computer entries are
  postponed until they are free
• Cashiers operate on the policy of
  First-Come-First-Served when serving both
  walk-ins and drive-ups
• With Water Dept, an additional cashier is
  available, and all cashiers handle all types of
  customers

42
Simulation Experiments

• Results analyzed over 30 independent days of
  operation
• Simulation programs written in SIMAN and
  animations developed in ARENA
• Animations shown for each alternative over 1 day
  (8 hours) of operation
• Over to animations .....

43
Average Time in System
44
Maximum Times in System
45
Maximum Drive-up Queue
46
Maximum Cashier Queue
47
Customers in Building
48
Average Server Utilizations
49
Maximum Computer Back-up
50
Methodology Summary

• Representation
• Step 1.0 Define Customer Classes
• Step 2.0 Define Routing vectors
• Step 3.0 Define Distance and Flow Matrices
• Analysis
• Step 4.0 Topological Diagrams
• Step 5.0 Layout Alternatives
• Step 6.0 Flow Analysis
• Synthesis
• Step 7.0 Evaluation of Alternatives
• Step 8.0 Synthesis
• Step 9.0 Implementation

51
Summary and Conclusions

• Dynamic Pedestrian Flows
• Fundamental Principles
• Travel Speed vs. Density
• C 5LW
• Case Studies
• System Optimization