CT-RAMP System Overview ¶

What is Activity-Based Modeling? ¶

Activity-based modeling represents a fundamental shift from traditional trip-based approaches to understanding travel behavior. Rather than predicting trips directly, activity-based models recognize that travel is derived from the need to participate in activities distributed across time and space.

Key Concepts ¶

Individual Decision-Making : Models individual persons and their daily activity patterns rather than aggregate zone-to-zone flows

Household Interactions : Recognizes that household members coordinate their activities and share resources (vehicles, childcare, etc.)

Time-Space Constraints : Explicitly considers the temporal and spatial constraints that shape activity participation

Behavioral Realism : Incorporates detailed behavioral theories about how people make activity and travel decisions

CT-RAMP System Architecture ¶

CT-RAMP implements activity-based modeling through a comprehensive system of interconnected choice models that simulate the daily activity and travel patterns of each person in a synthetic population.

System Components ¶

graph TB
    subgraph 1["Population & Context"]
        direction LR
        1a[Synthetic Population]
        1b[Land Use Data]
        1c[Transportation Networks]
        1a ~~~ 1b ~~~ 1c
    end

   1 --> 2
    subgraph 2["Long-term Choices"]
        direction LR
        2a[Auto Ownership]
        2b[Work & School Location]
        2a ~~~ 2b
    end

    2 --> 3
    subgraph 3["Mobility"]
        direction LR
        3a[Auto Ownership]
        3b[Transponder Choice]
        3c[Free Parking]
        3a ~~~ 3b ~~~ 3c
    end

    3 --> 4
    subgraph 4["Daily Activity Planning"]

        direction TB
        subgraph 4a["Coordinated Daily Activity Pattern"]
            direction LR
            4a1[CDAP]
            4a2[Explicit Telecommute]
            4a1 --> 4a2
        end

        4b[Mandatory]
        4c[Non-Mandatory]
        4d[Home]

        subgraph 4f["Individual Mandatory Tours"]
            direction TB
            4f1[Frequency]
            4f2[Time of Day]
            4f3[Tour Mode]
            4f1 --> 4f2
            4f2 --> 4f3
        end
        subgraph 4g["Joint Non-Mandatory Tours"]
            direction TB
            4g1[Frequency/Composition]
            4g2[Participation]
            4g3[Destination]
            4g4[Time of Tour]
            4g5[Tour Mode]
            4g1 --> 4g2
            4g2 --> 4g3
            4g3 --> 4g4
            4g4 --> 4g5
        end
        subgraph 4h["Individual Non-Mandatory Tours"]
            direction TB
            4h1[Frequency]
            4h2[Destination]
            4h3[Time of Day]
            4h4[Tour Mode]
            4h1 --> 4h2
            4h2 --> 4h3
            4h3 --> 4h4
        end
        subgraph 4i["At-Work Subtours"]
            direction TB
            4i1[Frequency]
            4i2[Destination]
            4i3[Time of Day]
            4i4[Tour Mode]
            4i1 --> 4i2
            4i2 --> 4i3
            4i3 --> 4i4
        end

        4a --> 4b
        4a --> 4c
        4a --> 4d
        4b --> 4f
        4f --> 4i
        4c --> 4g
        4c --> 4h

    end

    4 --> 5
    subgraph 5["Stop-level Choices"]
        direction LR
        5a[Stop Frequency]
        5b[Stop Location]
        5a --> 5b
    end

    5 --> 6
    subgraph 6["Trip-level Choices"]
        direction LR
        6a[Trip mode]
        6b[Auto Parking]
        6a --> 6b
    end


    classDef Household fill:#e8f5e8
    class 2a,2b,4a1,4g1 Household

Model Hierarchy and Flow ¶

CT-RAMP follows a logical hierarchy from long-term decisions to immediate travel choices:

1. Long-term Choices ¶

Auto Ownership Model - Determines household vehicle availability - Influences all subsequent mobility decisions - Considers household size, income, accessibility

2. Daily Activity Pattern Formation ¶

Coordinated Daily Activity Pattern (CDAP) - Coordinates activity patterns among household members - Determines mandatory vs. non-mandatory activity participation - Resolves household resource conflicts

3. Tour Generation ¶

Tours represent round trips from home for specific purposes:

Mandatory Tours - Work and school tours (required activities) - Generated based on person type and obligations

Joint Tours
- Multi-person household tours - Shopping, recreation, social activities

Individual Non-Mandatory Tours - Personal discretionary activities - Conditional on household coordination results

4. Tour Characteristics ¶

For each generated tour, the system determines:

Destination Choice - Where to conduct the primary activity - Based on accessibility, land use, competition

Mode Choice - How to travel (auto, transit, walk, bike, etc.) - Considers tour characteristics and level-of-service

Time-of-Day Choice - When to depart and return - Balances activity duration with travel conditions

5. Intermediate Stops ¶

Stop Frequency - How many stops to make en route - Separate models for outbound and return directions

Stop Location - Where to make intermediate stops - Considers stop purpose and routing constraints

Trip Mode Choice - Mode for each trip segment - May differ from primary tour mode

Behavioral Foundations ¶

Choice Modeling Framework ¶

CT-RAMP uses discrete choice models based on random utility theory:

Utility Functions : Mathematical representations of decision-maker preferences combining observed attributes and random variation

Logit Models : Probabilistic choice models that translate utilities into choice probabilities

Nested Structures : Hierarchical choice structures that capture correlations between similar alternatives

Coordination Mechanisms ¶

Household Resource Allocation - Vehicle assignment and sharing - Childcare and escort responsibilities - Joint activity participation

Temporal Coordination - Synchronized departure times - Shared ride arrangements
- Activity duration coordination

Spatial Coordination - Destination clustering for efficiency - Escort tour routing - Joint activity locations

Data Requirements and Scale ¶

Input Data Requirements ¶

CT-RAMP requires detailed input data to support microsimulation:

Synthetic Population - Individual persons with demographics, employment, school enrollment - Households with income, size, vehicle ownership constraints - Typically 2.5+ million persons in 1+ million households for Bay Area

Land Use Data - Employment by industry and occupation category - Households and population by income and demographic segments - Commercial floor space and attractions - Detailed geographic representation (40,000+ micro-zones)

Transportation Networks - All-streets highway network with traffic controls - Complete transit system with routes, schedules, fares - Pedestrian and bicycle facilities - Level-of-service matrices (travel times, costs) by time period

Computational Scale ¶

Modern CT-RAMP implementations handle:

Population Size - 1+ million households, 2.5+ million persons - 10+ million tours, 25+ million trips annually

Geographic Detail
- 40,000+ micro-analysis zones (MAZs) - 1,500+ traffic analysis zones (TAZs) - 3,000+ transit access points (TAPs)

Temporal Resolution - Half-hourly time periods (48 periods per day) - Peak spreading and congestion feedback

Model Innovation and Enhancements ¶

Advanced Features ¶

Transit Capacity and Crowding - Explicit modeling of transit vehicle capacity - Crowding discomfort in mode choice - Dynamic capacity constraints

Detailed Spatial Resolution - Micro-zone level destination choice - Walk access to specific transit stops - Realistic pedestrian routing

Emerging Mobility Options - Transportation Network Company (TNC) services - Ride-sharing and ride-hailing - Autonomous vehicle scenarios

Validation and Calibration ¶

Observed Data Integration - California Household Travel Survey (CHTS) - On-board transit surveys
- Census commute data - Regional travel behavior studies

Performance Metrics - Trip generation rates by purpose and person type - Mode share validation by geography and demographics - Temporal distribution and peak spreading - Accessibility and equity measures

Benefits and Applications ¶

Planning Applications ¶

Scenario Analysis - Land use and development patterns - Transportation infrastructure investments - Policy interventions (pricing, regulations)

Performance Evaluation - Accessibility impacts - Environmental outcomes - Social equity analysis - Economic development effects

Technical Advantages ¶

Behavioral Realism - Captures complex decision-making processes - Represents household interactions - Models activity-travel trade-offs

Policy Sensitivity - Responds to detailed policy variables - Captures distributional effects - Models emerging mobility options

Forecasting Capability - Robust response to demographic changes - Adaptive to new transportation technologies - Consistent with activity-based behavior

Implementation Considerations ¶

Computational Requirements ¶

Processing Time - Full population model: 8-24 hours on modern hardware - Parallelization opportunities at household level - Memory requirements: 32+ GB RAM for full implementation

Data Management - Large input datasets (10+ GB) - Detailed output files (5+ GB per iteration) - Quality assurance and validation procedures

Model Development Process ¶

Calibration and Validation - Parameter estimation using observed data - Sensitivity testing and reasonableness checks
- Iterative refinement and improvement

Quality Assurance - Extensive validation against observed patterns - Cross-validation with independent datasets - Sensitivity analysis and robustness testing

Integration with Regional Modeling ¶

CT-RAMP operates as part of an integrated regional modeling system:

UrbanSim Integration - Land use forecasts provide demographic and employment inputs - Feedback from accessibility to land development

Network Assignment - Trip matrices from CT-RAMP loaded onto transportation networks - Congested travel times feed back to accessibility calculations

Economic Modeling - Commercial vehicle models for freight and service trips - Economic impact analysis of transportation investments

Learning Resources ¶

For New Users ¶

1. Conceptual Foundation: Understanding activity-based modeling principles
2. System Architecture: How CT-RAMP components work together
3. Data Requirements: Input preparation and quality standards
4. Interpretation: Understanding and using model outputs

For Technical Users ¶

1. Model Specifications: Detailed utility functions and parameters
2. Calibration Procedures: Parameter estimation and validation methods
3. Customization: Adapting models for local conditions
4. Performance Optimization: Computational efficiency improvements

For Policy Analysts ¶

1. Scenario Development: Creating meaningful policy alternatives
2. Output Analysis: Interpreting results for decision-making
3. Uncertainty Assessment: Understanding model limitations
4. Communication: Presenting results to stakeholders

Last updated: September 26, 2025