Understanding and predicting pedestrian flows across a city is critical for smarter urban planning, infrastructure investment, and business decision-making. Traditional data collection is expensive, fragmented, and limited in scope, making large-scale insights difficult to obtain.

Our solution: a city-scale Agent-Based Modeling (ABM) platform that combines high-resolution spatio-temporal and GPS-based walking data with advanced Machine Learning to simulate and predict pedestrian behavior in diverse urban environments. Validated in Tel Aviv and 5 more cities in Israel, the model offers unprecedented accuracy in mapping how people move through streets, neighborhoods, and entire urban regions.

THE GAP

Current agent-based models of pedestrian flow focus on small-scale, indoor, or event-specific environments (airports, malls, stadiums). They do not deliver reliable, high-resolution predictions for full neighborhoods or cities—leaving a major gap for city planners, businesses, and risk assessors.

OUR APPROACH

• Data-driven behavior rules: Derived from GPS walking traces, travel surveys, and urban morphology.
• High-resolution simulations: From neighborhood level to entire cities.
• Robust calibration: Tested across areas with different physical layouts, functions, and demographics.

POTENTIAL MARKETS & USE CASES

1. Urban Planning & Transportation Authorities
   • Identify infrastructure gaps (sidewalks, crossings, shaded paths).
   • Model the effects of new transit lines, land use changes, or density increases.
2. Insurance Companies
   • Assess risk exposure to predict the likelihood of burglaries to buildings and cars in specific streets or zones.
   • Incorporate real mobility patterns into urban security and health risk models.
3. Commercial Retail
   • Evaluate foot traffic potential for new store locations, including targeted population (socio-economic targeted) and more.
   • Compare multiple sites based on predicted pedestrian patterns at different times of day/week.
4. Billboard & Out-of-Home Advertising Firms
   • Measure actual pedestrian exposure to advertising assets.
   • Price billboard space dynamically based on modeled attractiveness and visibility.
5. Event Planners & Emergency Services
   • Simulate pedestrian crowd flows during large or extreme events to optimize safety and accessibility.

STATUS

• Proof of Concept validated in 6 Cities including Tel Aviv with high predictive accuracy.
• Ongoing pilots in additional cities, including cities outside Israel.

WHY NOW

The urban environment is becoming denser, more dynamic, and data-rich. Pedestrian movement is essential and central in the current sustainability-oriented urban planning. Moreover, the ability to model and predict how pedestrians interact with the built environment is no longer just a planning tool—it’s an asset for insurance, retail, advertising, and risk management industries.

Further details are available upon request