Volvo Summer Workshop Track 2: Urban Transportation Physics: Volvo Summer Workshop Track 2: Urban Transportation Physics Carlos Daganzo


Outline: Outline Focus Mass motions in cities How things work? New technologies New policies Projects BLIPS Green Logistics Gridlock Self synchronizing buses Evacuation Safety


BLIP Concept: BLIP Concept Right lane reserved for bus but open to traffic when bus is not near by, like IBL. CMS and in-pavement lights dynamically restrict and allow access to lane. One CMS per block Rules of the road: Drivers bound by CMS message until next CMS is reached. Eichler and Daganzo, 2005, Bus Lanes with Intermittent Priority: Assessment and Design


Complexity Fades Away: Complexity Fades Away Results do not rely on particulars of how the “cocoon” is achieved. One example of cocoon formation.


Green Logistics – Implemented schemes: Green Logistics – Implemented schemes Geroliminis and Daganzo (2005), A review of green logistics schemes used in cities around the world


Water use - Coordinated transport : Water use - Coordinated transport A DHL-boat sails through the canals and serves as base-centre for bicycle-couriers reduction of 150.000 van-km / year TOTAL COST: 7,000€ Amsterdam - Floating Distribution Centre


Clean vehicles : Clean vehicles hybrid (clean and quiet) and energy efficient electric vehicles urban distribution centre (UDC) large trucks for long-distance transport to and from the UDC and of vans and small trucks for the centre. Total Cost: 1.2 million EURO combining an efficient goods distribution concept with the environmental impact of electric vehicle transport Rotterdam - ELectric vehicle CIty DIstribution System


Conclusions: Conclusions Promising city logistics schemes with “green” characteristics Largest and fastest growing cities in the developing world: ABSENT Schemes can be combined, adapted and modified to be of use (HOW? Research is necessary)


Gridlock In Cities: Gridlock In Cities Nikolas Geroliminis


Self-Stabilizing Bus Routes: Self-Stabilizing Bus Routes Josh Pilachowski


Statement of Problem: Statement of Problem Two parts of travel time Time to travel between stops Time to pick up and drop off passengers Bus routes by their very nature are unstable Demand is stochastic Small variation in demand can create large instability over time without control Queue proportional to headway Increased demand Slows bus Larger headway Increased demand Decreased demand Speeds up bus Smaller headway Decreased demand Buses eventually come together and act as a single unit


Sustainability: Sustainability Attracting Ridership Minimum ridership needed to derive environmental benefits Ridership from necessity as opposed to desire Spare-the-Air days Inherent Instability Routes with smaller headways are more unstable Smaller headways comes from higher use Equity Bus vs. Rail


Control Methods: Control Methods Station Control Holding Station Skipping Interstation Control Speed Control* Traffic Signal Preemption Other Adding/Removing Vehicles Early Turns


Existing Controls: Existing Controls Instability still exists Not many controls to alleviate this Establish slack time Estimate delays on route and add a fixed amount to the expected trip time Control points Determine certain stops as control points If a bus arrives at a control point ahead of schedule it waits there


Existing Controls: Existing Controls Expand on the idea of control points By increasing the number of control points errors have less time to propagate Control points require slack time built into the schedule A bus behind schedule cannot make up lost time easily Passengers don’t like sitting on the side of the road Station Control  Interstation Control Continuous control points become speed control


Speed Control: Speed Control Dynamic stabilization AVL allows buses to continually locate adjacent buses Replace slack time with a lower cruising velocity Allows buses to speed up to make up lost time Buses can slow down if they get too far ahead Trade off travel time for stability Lower average velocity (maybe) but lower variance Users have a higher cost for waiting time than for travel time


Possible Methods of Stabilization: Possible Methods of Stabilization Spring based stabilization Springs connecting adjacent buses k = spring constant ‘force’ based on spacing or headway F = k(hi – hi+1) v = vt + k Dt hi – k Dt hi+1 Unscheduled stabilized headways


Model Description: Model Description State Space: yi(t) - spacing hi(t) - headway Variables l - demand b - loss time per pax vt - target speed E[vi(t)] = vt/(1+vtlbhi(t)) Control: vt replaced by vi(hi(t),hi+1(t), yi(t),yi+1(t))


Testing the Model: Testing the Model Simulation Simple simulation Moves buses according to calculated velocity Picks up and drops off pax Real Data Faux bus route with students as ‘bus drivers’ Real bus route in Gothenburg, Sweden


Future Work: Future Work Additional areas of analysis and development Traffic Bus velocity constrained by traffic Parallel Bus routes Model interaction between overlapping routes? Create platoons? TSP/stoplights Handicap passengers


Logistics of City Evacuation: Logistics of City Evacuation Volvo Summer Workshop July 24, 2006 Stella So


Immobile Car-less; Challenged Car-owners: Immobile Car-less; Challenged Car-owners New Orleans, LA Houston, TX 3 weeks later…


Minimize { “evacuation time” }: Minimize { “evacuation time” } Increase capacity Contra flow Alternate routes Bottleneck clearance Manage demand Multi-modal ↑ pax per car ↓ “shadow” evacuation


Houston – a disaster for car-owners: Houston – a disaster for car-owners A Cut-Set Analysis… 23 h 19 h 6 h 8 h


We could’ve done MUCH better!: We could’ve done MUCH better!


Road User Adaptation to Road Safety Measures: Road User Adaptation to Road Safety Measures Presented by: Offer Grembek


Road Safety and Road Safety Measures: Road Safety and Road Safety Measures No Danger Danger Road Safety Generic risk factors Physical / Behavioral


Road Safety and Road Safety Measures: Road Safety and Road Safety Measures No Danger Danger Road Safety Road Safety Measures (RSM) Generic risk factors Physical / Behavioral “Engineered effect”


Road Safety and Road Safety Measures: Road Safety and Road Safety Measures No Danger Danger Road Safety Road Safety Measures (RSM) Generic risk factors Physical / Behavioral “Engineered effect” “Behavioral adaptation”


Behavioral Adaptation: Behavioral Adaptation What is behavioral adaptation Seatbelts All-red interval Current theories about behavioral adaptation Risk compensation theories (Wilde) Potential Benefits Road safety measures design and evaluation Academic


Research Question: Research Question Is the effectiveness of RSM influenced by an identifiable road-user behavioral adaptation? Are there specific types of RSM that generate behavioral adaptation? Are there specific conditions that generate behavioral adaptation? Do these behavioral adaptations have a significant impact on the effectiveness of RSM? How does the impact of these behavioral adaptations change over time?


Road Safety and Behavioral Adaptation: Road Safety and Behavioral Adaptation Inconsistencies in RSM and behavioral adaptation Mandatory seat-belts Center high mounted stoplights Setbacks of using RSM studies Statistical validity (confounding, regression-to-the-mean) Insufficient data


The Potential of a Different Approach: The Potential of a Different Approach Industrial safety and security Detailed longitudinal data Immediate consequences Tradeoff between safety and productivity Epidemiology Heterogeneous study population No visible benefits of compliance Non-compliance due to thrill (safe-sex) Portfolio theory Human behavior under risk


Future Work: Future Work Carlos Daganzo


Future Work – Data Needs: Future Work – Data Needs


Transit/City Structure: Transit/City Structure Buses - Small


Transit/City Structure: Transit/City Structure Buses - Medium


Transit/City Structure: Transit/City Structure Rail - Small


Transit/City Structure: Transit/City Structure Rail - Large