CIVE 341: Transportation Engineering Applications
Estimated study time: 9 minutes
Table of contents
Sources and References
Primary texts — Roess, R.P., Prassas, E.S., and McShane, W.R., Traffic Engineering, 5th ed., Pearson, 2018; Transportation Research Board, Highway Capacity Manual (HCM) 6th Edition, 2016 (public summaries).
Supplementary texts — Garber, N.J. and Hoel, L.A., Traffic and Highway Engineering, 5th ed., Cengage, 2015; Mannering, F.L. and Washburn, S.S., Principles of Highway Engineering and Traffic Analysis, 7th ed., Wiley, 2020.
Online resources — MIT OCW 1.225J “Transportation Flow Systems”; FHWA Signal Timing Manual; Transport Canada technical reports; TRB open research summaries; TAC Geometric Design Guide public summaries.
Chapter 1: Traffic Engineering Foundations
1.1 Traffic Stream Parameters
Flow \( q \) [veh/h], density \( k \) [veh/km], speed \( v \) [km/h] related by \( q = kv \). Types of averaging: time-mean speed (average of point observations) > space-mean speed (instantaneous average in a distance). Headway (time between successive vehicles) and spacing (distance) are microscopic analogs.
1.2 Fundamental Diagram
\( q \)-\( k \) relationship shows flow increases with density to capacity, then decreases. Greenshields linear model: \( v = v_f(1 - k/k_j) \), \( q = v_f k(1 - k/k_j) \), maximum flow \( q_c = v_f k_j / 4 \). Nonlinear models (Underwood, May, triangular) refine fit. Capacity drop (5-15%) from free-flow to queue-discharge explains freeway breakdown persistence.
1.3 Measurement
Loop detectors (inductive, measure volume and occupancy), radar, lidar, cameras with computer vision, cellular/Bluetooth probes, floating car runs. Each technology has characteristic error and cost profile. Traffic counts (short-term, 24-hour, 48-hour, continuous) sampling strategies balance precision with resource.
1.4 Volume, Design, and Level of Service
Annual average daily traffic (AADT), design hourly volume (DHV, typically 8-12% of AADT), peak hour factor (PHF). Volume/capacity ratio and level of service (LOS, A-F) applied to freeways, multilane highways, two-lane highways, urban streets, intersections, weaving sections, ramps. HCM methods differ by facility type.
Chapter 2: Travel Forecasting
2.1 Demand Forecasting Frameworks
Four-step models (generation, distribution, mode choice, assignment) remain dominant in practice. Activity-based models offer deeper behavioral representation; agent-based simulation further resolves individual decisions. Forecasting horizons: 5-10 years (operational), 20-30 years (strategic planning).
2.2 Traffic Assignment
User equilibrium (Wardrop 1st): no user can reduce travel time by unilaterally changing routes. System optimal (2nd): total travel time minimized. Real systems operate near UE; SO requires centralized control or pricing. Stochastic user equilibrium accounts for perception errors.
BPR link performance function:
\[ t(x) = t_0\!\left[1 + \alpha(x/c)^\beta\right], \]with free-flow time \( t_0 \), capacity \( c \), typical \( \alpha = 0.15 \), \( \beta = 4 \). Frank-Wolfe and MSA algorithms solve UE on large networks.
2.3 Demand Responsive Effects
Induced demand: new road capacity generates new traffic; elasticity ~0.6-1.0 over long run. Changes to access time reshape mode split, destination choice, and eventually land use. Neglecting induced demand systematically overstates benefits of road expansion.
2.4 Scenarios and Uncertainty
Base, high, low growth scenarios; technology scenarios (autonomous vehicles, electrification); policy scenarios. Monte Carlo analysis propagates uncertainty in demographics, economy, fuel prices. Robust decision-making prefers options that perform reasonably across scenarios over those optimized to a single expected future.
Chapter 3: Intersection Analysis
3.1 Signalized Intersection Capacity
Saturation flow rate \( s \) [veh/h-green] from ideal \( s_0 \) (typically 1900) with adjustments for lane width, grade, heavy vehicles, parking, bus blockage, local adjustment factor. Capacity of a movement:
\[ c = s \cdot g/C, \]with effective green \( g \), cycle length \( C \). Volume/capacity \( v/c \) ratio drives delay and LOS.
3.2 Signal Timing
Webster’s optimum cycle length:
\[ C_o = \frac{1.5 L + 5}{1 - \sum Y_i}, \]with lost time \( L \), critical flow ratios \( Y_i = v_i/s_i \). Green splits proportional to critical \( Y_i \). Delay formulas (Webster, HCM incremental) combine uniform and random arrival components.
3.3 Coordination
Through a corridor, offsets between adjacent signals can produce “green waves” for the major direction of travel. Time-space diagrams visualize platoon progression. Bandwidth-based (MAXBAND) and delay-based (TRANSYT) optimization methods. Adaptive systems (SCOOT, SCATS) adjust timings in real time based on detection.
3.4 Alternative Intersections
Roundabouts (reduce speed, simplify conflicts, eliminate severe angle crashes), displaced left turn (DLT), restricted crossing U-turn (RCUT), continuous flow intersection (CFI), diverging diamond interchange (DDI). Each has specific demand and geometry niches.
Chapter 4: Operational Control and ITS
4.1 Ramp Metering
Fixed-time, local-traffic-responsive (ALINEA), coordinated freeway-wide algorithms. Benefits: delays breakdown, smooths merges, increases mainline throughput. Queues at ramps and fairness (meter bypass equity) are implementation concerns.
4.2 Dynamic Lane Management
High-occupancy vehicle (HOV), high-occupancy toll (HOT), dynamic pricing, managed lanes. Responsive speed limits (VSL) during congestion smooth flow and reduce rear-end crashes. Part-time shoulder use (hard shoulder running) as temporary capacity.
4.3 Traveler Information
Dynamic message signs (DMS), 511 phone and web, mobile apps, in-vehicle navigation. Accuracy and timeliness of information drive user response. Some systems actively shape routing; others report to empower choice.
4.4 Connected and Automated Vehicles
V2V and V2I communications (DSRC, C-V2X) enable cooperative adaptive cruise control, signal phase and timing messages, eco-driving. Fully automated vehicles may shift demand, crash profiles, and infrastructure requirements; penetration timelines remain uncertain.
Chapter 5: Safety Analysis
5.1 Crash Data and Analysis
Crash reports (MV-104 equivalents), geocoded locations, severity (fatal, injury, property damage). Crash rate \( = \) crashes per million vehicle-miles / entries. Network screening identifies sites with excess crashes relative to predicted; empirical Bayes adjusts for regression-to-mean.
5.2 Highway Safety Manual Framework
HSM (AASHTO, 2010) provides:
- Safety Performance Functions (SPFs): predict crash frequency as function of AADT and geometry.
- Crash Modification Factors (CMFs): ratio of expected crashes with treatment to without.
- Economic analysis: crash costs (fatal ~$11 million US, injury $600k, PDO $10k) against treatment cost.
5.3 Countermeasures
Proven engineering treatments: centerline and shoulder rumble strips, skid-resistant surfaces, improved delineation, roundabouts, median barriers, protected-only left turns, pedestrian refuge islands, leading pedestrian intervals. Systemic approach targets treatment locations by risk rather than only high-crash sites.
5.4 Vulnerable Road Users
Pedestrian and cyclist fatalities overrepresent all transport deaths. Safe system approach: reduce speeds (30-40 km/h in residential zones), separate modes where speeds exceed survivability thresholds, design for human error and tolerable consequences.
Chapter 6: Economic Evaluation
6.1 Benefit Categories
Travel time savings, vehicle operating costs, safety (crash reduction), emissions (air quality, climate), noise, construction cost, maintenance. Values of time (VOT) derived from stated/revealed preference; typically $15-25/h for commuting in North America.
6.2 Cost-Benefit Metrics
Net present value:
\[ NPV = \sum_{t=0}^{T}\frac{B_t - C_t}{(1+r)^t}. \]Benefit-cost ratio, internal rate of return, equivalent uniform annual benefit. Discount rate (3-7% typical public projects) dominates long-horizon outcomes; social discount rate lower than private reflects intergenerational equity.
6.3 Multi-Criteria Analysis
Non-monetizable attributes (equity, aesthetics, strategic) handled through scoring and weighted sums, or outranking methods (ELECTRE, PROMETHEE). Sensitivity to weights and rating uncertainties essential.
6.4 Life-Cycle and Externality Costing
Infrastructure life-cycle cost: capital + operations & maintenance + rehab + end-of-life. Internalize externalities through shadow pricing or explicit environmental and social accounts. LCCA (Life-Cycle Cost Analysis) standard in pavement type selection, bridge design, and asset management.
