The team is developing a system model designed to minimize surface transportation energy consumption using distributed simulation modeling, big data analytics, and agent interaction.
The basic conceptual approach includes:
- Monitoring system-wide, real-time, high-resolution traffic conditions (big data applications)
- Applying statistical methods to map behavioral responses to system perturbations over time and create a space-time memory
- Linking big data analysis and pattern recognition with simulation
- Estimating energy consumption, time, and cost tradeoffs for departure times, modes, and routes
- Delivering tailored messages to participants to support travel planning and demand-responsive decision making
The data-driven system will employ: 1) real-time, 20-second, lanespecific operations data from regional intelligent transportation systems; 2) revealed origin-destination patterns and arterial speeds from AirSage cell phone tower data; and 3) second-by-second position/speed data from 40,000 volunteers via Smartphone app.
Instead of focusing on large, centralized simulations, the team will implement more agile, distributed, real-time simulations, which will ultimately reside on network agent computers (e.g., signal control cabinets) and agent devices (vehicle/smart phone), to predict nearfuture, corridor-level traffic conditions.
Instead of focusing on minimizing total energy use, in which the travel times of some travelers increase so that the travel times of many travelers can be reduced slightly, the team is focused on individual optimization. The system will be designed to reduce energy consumption and emissions from individual agents, without increasing the agent’s travel cost and origin-destination travel time (where previous repeat travel observations serve as the baseline).