How can you predict travel demand?

1 Data Collection

The first step of travel demand forecasting is to collect and analyze data on the existing and future characteristics of the transportation system and the travelers. This includes information on the network, such as road capacity, speed, and congestion; the modes, such as transit, bike, and car; the land use, such as population, employment, and zoning; and the travel behavior, such as trip frequency, purpose, mode choice, and route choice. Data can be obtained from various sources, such as surveys, censuses, traffic counts, GPS, and smart cards.

2 Model Development

The second step of travel demand forecasting is to develop mathematical models that represent the relationships between the variables that affect travel demand. These models are usually based on the four-step model, which consists of trip generation, trip distribution, mode choice, and trip assignment. Trip generation estimates the number of trips produced and attracted by each zone; trip distribution allocates these trips between origin and destination zones; mode choice determines the mode of travel for each trip; and trip assignment assigns each trip to a specific route on the network. Each step can use different methods, such as regression, gravity, logit, or network analysis.

3 Scenario Analysis

The third step of travel demand forecasting is to apply the models to different scenarios of interest, such as changes in network supply, mode availability, land use patterns, travel costs, or policies. For each scenario, the models can estimate the changes in travel demand, such as trip volume, mode share, vehicle miles traveled, or traffic flow. These outputs can then be compared with the base case or other scenarios to evaluate the effects of the proposed alternatives.

4 Model Validation

The fourth step of travel demand forecasting is to validate the models and the outputs by comparing them with observed or expected data. This can help to check the accuracy, reliability, and sensitivity of the models, and to identify any errors or biases. Validation can be done at different levels, such as aggregate, disaggregate, or behavioral. Some of the common validation measures are goodness-of-fit, error statistics, elasticity, or calibration factors.

5 Model Application

The fifth step of travel demand forecasting is to use the models and the outputs for transportation planning purposes, such as identifying problems, generating solutions, assessing impacts, or prioritizing investments. The models and the outputs can provide useful information and insights for decision-making, such as demand-supply gaps, mode shifts, congestion levels, emissions, or benefits and costs. However, the models and the outputs should be interpreted with caution, as they are based on assumptions, simplifications, and uncertainties.

6 Model Improvement

The sixth step of travel demand forecasting is to improve the models and the outputs by incorporating new data, methods, or technologies. Travel demand forecasting is a dynamic and evolving process, as the transportation system and the travelers change over time. Therefore, the models and the outputs should be updated and refined regularly to reflect the current and future conditions and preferences. Some of the possible ways to improve the models and the outputs are to use more disaggregate data, to capture more behavioral factors, to account for more modes and interactions, or to adopt more advanced techniques, such as agent-based or activity-based modeling.

7 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?