An interactive web platform (and Control Tower component) for managing, predicting, optimizing, and analyzing transport fleet arrivals into a geographic destination (e.g., a gate). It enables fleet managers to configure arrival windows, routes, speeds, and travel conditions, compare robust and flexible scheduling strategies, and stress-test performance under traffic and weather disruptions using Monte Carlo simulations. Users may use a mobile app to access the recommend optimized results.
-
Main Gain(s): Reduction of CO2 emissions….
-
Use Case(s): Event data generated from NEXUS partners historical data (no real-time data was provided)
-
Start TRL: 0 - (evidences)
-
Final TRL: 0 - (evidences)
-
Main Contributions: (theoretics) A. Iglesias, E. Rocha; (implementation) A. Iglesias; (integration) E. Rocha
Main Features
- Interactive configuration sidebar for arrival windows, routes, speeds, and travel times
- Real-time optimization results with charts, maps, and KPIs
- Comparison between robust and flexible scheduling approaches
- Monte Carlo simulation for traffic and weather disruption testing
- Interactive tables, maps, and Gantt charts for operational analysis
- Backend optimization using numerical solvers and stochastic simulations
- Asynchronous computation to maintain smooth user experience
- Can receive real-time data and send back the recommend transport speeds, arrival times, etc.
Videos
| Web Appplication | Public Demonstration on a Control Tower |
Product Development Context
This software platform is developed within the strategic scope of the NEXUS Project, specifically contributing to the architecture of an intelligent Logistics Control Tower. The Control Tower serves as a centralized hub designed to support end-to-end supply chain visibility, coordination, and real-time decision-making. Within this framework, managing and synchronizing inbound and outbound freight flows is critical to mitigating bottlenecks. This product bridges the gap between high-level Control Tower visibility and operational execution by introducing a dedicated module for truck arrival scheduling optimization. By integrating real-time mathematical programming with stochastic simulation, this platform provides the Control Tower with the predictive and prescriptive capabilities necessary to dynamically orchestrate fleet movements, absorb disruptive regional anomalies, such as weather or traffic, and maintain seamless synchronization across the entire logistics network.
Product Definition and Benefits for Users
Over the past few months, we have designed and implemented an interactive web platform conceived as the central interface for managing, visualizing, and analyzing our truck arrival scheduling optimization model. Our platform serves as an interactive bridge between complex mathematical algorithms and the fleet manager. Through it, users can input key parameters such as prescribed arrival windows, routes, and initial fleet conditions to instantly trigger the optimization algorithms. The application renders optimization results in real time using dynamic charts, maps, and key performance indicators, or KPIs. This enables immediate comparison between:
- The robust approach, which strictly maintains the original arrival sequence * The flexible approach, which allows fleet reordering in the event of critical delays Furthermore, the platform integrates the Monte Carlo simulation module, providing a visual environment where users can stress-test the system by introducing perturbations into regional penalty factors, such as traffic and weather, and instantly observe both models´ adaptability and resilience via probability curves and sensitivity analyses.
Product Characterization – Technical Specs
1. Frontend
The platform´s frontend was designed prioritizing usability, visual clarity, and interactive decision-making. The interface is structurally divided into:
- A configuration sidebar * A central results display panel * A graphical panel In the configuration section, the user introduces numerical inputs for time windows, route selection, initial speeds, and initial travel time. The main panel acts as an informative dashboard, which allows for visual verification of the prescribed speeds as well as the arrival times of the trucks after optimizing the model. In the third tab, interactive data tables and map components are incorporated to georeference regions with the highest penalties, alongside Gantt graphs that illustrate the results of the simulations. This setup ensures that abstract mathematical concepts, such as non-linear programming and uncertainty, are perfectly comprehensible and actionable for logistics operators.
2. Backend
The backend of the application constitutes the analytical and computational core of the system, developed entirely in. This layer hosts the non-linear optimization model responsible for solving the arrival time problem. To solve this constrained optimization problem, the backend utilizes numerical solvers, which process the input data matrices, road speed constraints, and regional weather and traffic penalty matrices. When the user triggers a Monte Carlo simulation, the backend efficiently executes hundreds of stochastic iterations, generating random perturbations based on predefined probability distributions for the regional penalties. All data flow control logic, fleet state management, delay calculations, and the reordering algorithm for the flexible strategy are processed asynchronously on the server. This ensures that heavy computations do not freeze the user experience, while delivering a clean presentation of the results.
Product Testing, Validation and Evaluation
This phase assessed the platform´s reliability and performance. Through systematic validation, the optimization outputs were verified and analyzed, as well as all the graphical tools.
Product’s Internal and External Limitations/Restrictions
The limitations of the visualization tool are directly extrapolated from the inherent weaknesses of the optimization models, creating a synergy between the two. That is, the more realistic our theoretical model becomes, the higher the quality of the results presented on our platform.