Network Thinking in Transport Systems Milos N Mladenovic

Network Thinking in Transport Systems Milos N. Mladenovic Assistant Professor Spatial Planning and Transportation Engineering Department of Built Environment 05. 10. 2016

Lecture Outline • • • Introduction – the origins of complexity Network thinking basics Network thinking – Example 1 Network thinking – Example 2 Network thinking – Example 3 Conclusions 17. 3. 2018 2

Literature • Vukan R. Vuchic, Urban Transit Operations, Planning and Economics, John Wiley and Sons, 2005 • Hamdy Taha, Operations Research: An Introduction (8 th edition), Pearson Prentice Hall, 2007 • Public transport – Planning the networks, Hi. Trans Best practice guide 2, 2005 • Jean-Paul Rodrigue, Geography of Transport Systems, 2013 • Levinson, David M. , and Kevin J. Krizek. Planning for place and plexus: metropolitan land use and transport. Routledge, 2007. • Schönfelder, Stefan, and Kay W. Axhausen. Urban rhythms and travel behaviour: spatial and temporal phenomena of daily travel. Ashgate Publishing, Ltd. , 2010. 17. 3. 2018 3

Evolution of the Spatial Structure of a City 17. 3. 2018 4

Transportation System Evolution with Urban Area Growth 17. 3. 2018 5

Scale and Urban Spatial Structure 17. 3. 2018 6

Dynamics of Urban Change 17. 3. 2018 7

The Foundations of Network Thinking 8

Complexity 1. Simple problems 2. Disorganized complex problems 3. Organized complex problems 1. and 2. seek equilibrium approach with reductionist, deterministic, and statistical methods 3. heterogeneity, coherent local interaction, irreducibility, persistent disequilibrium 17. 3. 2018 9

Network Thinking • A set of factors • Relationships (interdependencies) between factors • Emergent properties - the system is more than the sum of its elements because of the interaction Systemic perspective Policies Interface Infra. User Marketing Operation Pricing Factors in planning flexible transport service 10

Network Theory • Network theory: an introduction • Transport network as a set of nodes (vertex) and branches (edges/links) • There is transportation demand at certain nodes • How should a network of urban mass transport look like? • What should be a route for a freight vehicle delivering goods to the stores? • Everything is a network? • Node: person, web-page, company, stations • Links are interactions between nodes 11

Network Theory • First problem by Leonhard Euler (1707‑ 1783) • City of Königsberg, two river islands, seven bridges • Can a person cross, in one attempt, over all the seven bridges only once? • No, she cannot 12

Network Theory • Transportation network G = (N, A) where a set of nodes is N, and a set of links (edges) between the nodes is A • (i, j) is a link that connects node i N with node j N • Every link (i, j) A can have one or more numerical characteristics: cij – link cost; uij – link capacity; • Path is a set of links from node i to node j – e. g. , (1, 6, 7) or ((1, 6), (6, 7)) • Cycle starts and ends in the same node – e. g. , (1, 8, 7, 6, 1) 13

Network Theory Non-oriented network Oriented network Mixed network 14

Network Connectedness • “Well connected” metro line • Network connectedness level enables cross-comparison among networks and evolution of network over time • Planar graph – links intersect only in nodes v – number of nodes in the planar graph emax – maximum number of links Maximum connectivity of planar graphs 15

Network as a Matrix If there is a link between (i, j) else 16

Node Accessibility • Node accessibility A D E 0 1 1 0 0 A 2 B 1 0 1 1 0 B 3 C 1 1 0 1 1 C 4 D 8 C A 8 B 0 1 1 0 1 D 3 E 0 0 1 1 0 E 2 Node degree 17

Example 1 – Transit System Planning 18

Components of an Urban Transit System 19

Transit Plan Development Local conditions: - Topography Transit system goals and objectives: Demand: - Street network Role of transit - Forecasted travel volumes - Land use - Origin-destinations - Financial resources - Required level of service - Environment Determine generic mode: - Rapid transit - Semirapid transit Design alternative plans I, III, etc. with combinations of - Mode Evaluation criteria - Network alignments - Network length Evaluation and selection of alternative plans Final plan: Mode, network and lines 20

Two Contrasting Network Design Principles 21

Important Route and Network Attributes • Route lengths • Route pattern (radial, grid, composite, etc. ) • Temporal demand profile on each route (peaking factors by time of day) • Spatial demand profile (distribution along routes) • Operating speeds along routes • Frequency along routes • Frequency of connections • Stop or station spacing along routes • Area coverage • Transfer facilities – between public transport services • Transfer facilities – between public and other modes 22

Transit Line, Network, Station Assuming that each spacing between stations is 1 km long, the values in this network are: Line (route) lengths: LAC = 6 km, LAD = 5 km, LEF = 7 km. Network length: LAC + LBD + LEF = 15 km. Total line (route) length: LAC + LAD + LEF = 18 km. 23

Route Length • Advantages of long lines: • Long lines serve more trips directly than short lines • Long lines have smaller proportion of terminal time because technical maneuvers (changing ends of the train or turning vehicles) are almost independent of line length • Disadvantages of long lines: • Long lines may result in less efficient scheduling and run cutting because long cycle times are difficult to fit into shifts • Major delay propagation 24

Line Types Inner line in center Diametical line Radial line Circular line Orbital/tangential line Double circular line Feeder line Circular-radial line Further info http: //transitmap. net/ 25

Rapid Transit Network 26

Diametrical Lines • Transverse, direct, through (center) lines that connect suburbs on difference sides of the city center • Two sections should be balanced in terms of design passenger volume • No downtown terminal needed • Transfer of delay from inbound to outbound • L shaped lines are less direct but more transfer points and better area coverage 27

Radial and Branch Lines Radial • One terminus in center, other in the suburbs • Tend to follow major demand directions • Decreasing passenger demand outward from the center • Some TUs often operate only to intermediate points (short-turn service) Branch • Complements of radial and diagonal lines • Broader coverage in suburban areas • Street transit modes (e. g. , bus, streetcar) have limitation on the number of brunches due to congestion in the trunk section 28

Radial Lines - San Francisco 29

Network vs. City 30

Planning Strategy and Path Dependency 31

Planning Strategy and Path Dependency 32

Basic Network Characteristics City System Opening/ Latest Construction Network Length (km) Number of Stations Average Station Spacing (m) Maximum curve radius (m) Maximum gradient (%) Station platform length (m) Power voltage (V) 1. Chicago 1892/1983 142 1, 100 27 3. 5 100 -130 630 2. Hamburg 1912/1980 89. 5 80 1, 075 125(68) 5. 0 90/100/125 750 3. St. Petersburg 1955/1975 56 34 1, 440 n. a. 80 825 4. London 1863/2000 387. 6 249 1, 300 101 3. 5 130/109 600 5. Madrid 1919/2003 50. 9 84 550 90 5. 0 60/90 600 6. Mexico 1969/1980 40. 8 48 830 105 7. 0 150 7. Montreal 1966/2007 40 h 43 h 853 140 6. 3 152 750 8. Moscow 1935/1980 171 103 1, 550 250 4. 0 155 825 9. New York 1868/1979 385 477 805 52 3. 0 160 -188 600 10. Paris (Metro Urb. ) 1900/1994 178. 1 348 538 75(40) 4. 0 75/90/105 750 11. Philadelphia SEPTA 1907/1972 39. 4' 53 740 43 5. 0 105 - 118 600 12. Stockholm 1950/1978 102 72 h 1, 000 200 (120) 4. 0(4. 8) 145 650 13. Tokyo (TRTA) 1927/1980 127. 6 119 840/1, 660 91. 4. 0 120/220 14. Toronto 1954/1980 53 55 880 122 3. 5 152 600 15. West Berlin 1902/1980 92 109 761 74 4. 0 85 -110 780 600/1500 33

Example 2 – Travel Behavior 34

Principle agents influencing metropolitan patterns of the built environment Individual behavior (2) Home buying (3) Job seeking (4) Traveling on networks (5) Scheduling time (6) Firm Behavior (7) Government behavior(10) Developers Anticipate the needs of Locators Non-retail locating: setting a business (8) Retail Locating: Selling (9) Operating place and plexus (13) Assembling plexus (12) Designing place and plexus (11) 35

Home Buying and Job Seeking • Long-term choices affect origin and destination of human activities • Access to services and employment • Access to transport modes • Regional vs. neighborhood accessibility • Cultural and lifestyle factors, e. g. , aversion, self-image, house size, num. of bedrooms, fireplace, back yard, etc. • Game-oriented behavior • Social networks • Complex pay-off function 17. 3. 2018 36

Two Forces of Complexity • Risk aversion • Variety seeking • Sensation seeking as desire for unfamiliar • Alteration among familiar alternatives • Acquisition of new information 17. 3. 2018 37

Scheduling • Motivation to execute an activity • The choice of a potential options • The result of the decision-process • Access to mobility tools Space-time prism 17. 3. 2018 38

Individual (habitual) Decisions • Rhythmic structure of time use and travel is a fundamental element of daily life • Recurrent activities/trips can be observed in anyone’s life if observed over prolonged period • Daily or weekly rhythms that differ based on trip purpose • Continual public transport use, as car use, walking and biking tends to be a habitual behavior – lifestyle part (occupational status) • Habit = learned sequence of acts that have become automatic responses to specific cues, and are functional in obtaining certain goals or end-states • Daily decisions are instances of long-term behavior 17. 3. 2018 39

Individual (habitual) Decisions • The structure of daily destination choices is dominated by few locations • There is a permanent process of discovery of “new” location at constant innovation rate • The size of the activity space is relatively stable over longer periods • Travelers tend to spatially cluster their activities • Workplace is more isolated in terms of activity clustering than before • Spatial alternatives are often limited, clustered, and unevenly known • Habits affect assessment and evaluation of alternatives 17. 3. 2018 40

Social Context • Sum of the social signals which traveler sends and receives by participating in the particular trip or activity • Monetary aspects (wage earned, restaurant bill) • The size or the composition of the party with whom the travel or activity is undertaken • Prestige associated with certain location • Whether activity is a part of a long-range plan • Whether activity helps to keep a promise • Traveler’s household context • Etc. • Meaning of the activity and travel 17. 3. 2018 41

Social Practice Theory • Beyond Attitude, Behavior, Choice (individual choice) framework • Take practices rather than the individuals who carry them as core unit of analysis • ‘a routinized type of behavior which consists of several elements, interconnected to one another: forms of bodily activities, forms of mental activities, ‘things’ and their use, a background knowledge in the form of understanding, know-how, states of emotion and motivational knowledge (Reckwitz 2002) Competence Material Meaning 17. 3. 2018 42

What is the Y axis? Phone call activity over week 17. 3. 2018 43

Movement as a Network • Networks consist of various subnetworks, either spatially or temporally • Motif - subnetworks that occur more often than in the randomized versions of the entire network 17. 3. 2018 44

Clustering Movement Types 17. 3. 2018 45

Trip Share in Finland 17. 3. 2018 46

Trip Duration 17. 3. 2018 47

Trip Distance 17. 3. 2018 48

Example 3 – Alternative Plan Evaluation 49

Evaluation Typology • Evaluation prior to plan implementation • Evaluation of alternative plans • Analysis of planning documents • Evaluation of planning practice • Studies of planning behavior • Description of the impacts of planning and plans • Policy implementation analysis • Evaluation of the plan implementation 17. 08. 2015. 50

Comparison of Alternatives 51

Comparison of Alternatives • Multiple alternatives • Multiple criteria • Conflicting criteria http: //en. citizendium. org/ 52

Multi-Attribute Decision Making • Multi-Criteria Decision Making (MCDM) is decision theory approach and set of techniques that helps in coherent ordering of conflicting options • Multi-Attribute Decision Making (MADM) is used when decision space is discrete and when problem is semi-structured • Each MADM evaluation model is defined by: • a set of alternatives • a set of criteria or attributes for evaluation • a decision matrix 17. 08. 2015. 53

Simple Additive Weighting Example Alternative 1 Alternative 2 WEIGHT Do nothing Alternative 3 Metro Bridge Criteria 1 Criteria 2 Cost Safety 0. 2 90 35 20 90 50 40 Criteria 3 Emissions 0. 2 30 80 40 Criteria 4 Criteria 5 Reliability Land use 0. 2 1 40 50 49 60 35 57 50 40 44 17. 08. 2015. 54

Conclusions 55

Conclusions • Everything (complex) is a network! • Developing a better understanding of system functions and interactions • “We will interfere less, but in more appropriate ways” • Think about the activities for different individuals in different households in space and time • Think about relationships to other elements of living • Think about a wider context of the network • Think about wider context of planning and policy objectives 17. 08. 2015. 56

Thank you