Skip to main content

Research & Innovation

Advancing spatial intelligence through reproducible simulation benchmarks, academic research partnerships, and exploration of hybrid classical-quantum optimization algorithms.

🔬

Microscopic Traffic Simulation

Validating multi-agent reinforcement learning policies inside Eclipse SUMO. Simulating tens of thousands of individual vehicles across real-world geometries to measure queue dissipation, travel time variance, and emissions.

Active Experiments
🇪🇺

Horizon Europe Grant Consortia

Actively structuring research proposals targeting Horizon Europe (Cluster 5: Climate, Energy & Mobility) and EIT Urban Mobility calls for non-dilutive funding to validate territorial digital twins.

2026–2028 Pipeline
⚛️

Hybrid Classical-Quantum R&D

Investigating Quantum Approximate Optimization Algorithms (QAOA) and Quantum Reinforcement Learning on cloud simulators to model NP-hard combinatorial multi-modal evacuation routing.

Long-Term Concept

Simulation Transparency & Verification Standards

In enterprise transportation engineering, mathematical claims must be rigorously qualified. All performance metrics reported by MoviSabio at this stage reflect in-silico microscopic simulation benchmarks inside calibrated digital twins:

Simulation Setup (SUMO & TraCI)

Networks are extracted from OpenStreetMap and calibrated with Krauss car-following models. In synthetic benchmark corridors, multi-agent DQN agents demonstrated up to 45% vehicular delay reductions compared to uncoordinated fixed-cycle timings.

Field Deployment Boundary

Physical field actuation requires municipal authorization, controller cabinet safety compliance (NTCIP 1202), and shadow-mode calibration. Simulated results serve to justify scoping trials, not claim closed municipal deployments.

Academic & Lab Collaboration

We invite university transportation institutes, GIS research groups, and climate labs to co-author benchmarks.

Transportation Engineering Labs

Access our SUMO TraCI Python codebase to evaluate novel reinforcement learning reward functions and adaptive green-wave offset algorithms against benchmark networks.

Geospatial & Urban Planning Institutes

Collaborate on OpenStreetMap digital twin ingestion pipelines and spatial network vulnerability analysis during extreme weather events.

Environmental & Climate Centers

Co-develop micro-emissions calculation models validating the correlation between intersection queue reduction and localized PM₂.₅ dissipation.

Initiate Academic Collaboration →