DYNAMIC OPTIMIZATION
On AIR acquired a wide experience in developing dynamic optimization models, applied to decision support systems for industrial applications, traffic management and scheduling in complex logistic networks.
METHODS
Graph modelling and Shortest path algorithms, Dynamic programming algorithms, Conflicts detection and resolution, Evolutive Programming, Reinforcement Learning, Quantum Com- puting.
APPLICATIONS
Railway online scheduling, Ship voyage optimization, Production planning, Road traffic man- agement, Intermodal logistics, Port automation.
STATISTICAL MODELING
OnAIR has advanced skills in statistical modeling of complex systems, in order to infer predic- tive information about their behaviour.
METHODS
Unsupervised anomaly detection, Online learning of system behaviour, Advanced multivari- ate statistics, Spatial/Time series forecasting, Learning-based feature engineering, Cluster analysis and mapping.
APPLICATIONS
Digital twin design and fine tuning, Predictive maintenance of machineries, Decision support systems for industry/services, Drivers identification in ecology, Adaptive user profiling and grouping, Customized thematic maps for land and sea.
COMPUTER VISION
On AIR accumulated a relevant background in realizing software tools for many computer vision applications. On AIR operates as a partner for system integrators, designing and devel- oping software tools tailored to the user needs.
METHODS
2D and 3D shape recognition, OCR and symbol recognition, Underwater image processing, Multi-camera calibration, Object detection and motion tracking, Deep learning for image segmentation.
APPLICATIONS
Road traffic monitoring and analysis, Advance videosurveillance, Monitoring of underwater ecosystems, Remote sensing for agriculture and sea monitoring, Automatic visual inspection, Optical microscopy and cell counting, Human behaviour understanding.
ROBOTICS
OnAIR integrates its competence with sensor and embedded platforms to put intelligence on boards of several types of robotic systems, working in many different contexts
METHODS
Robot path planning, Path following by visual features, Multi-sensor data processing, Data integration and synchronization, Adaptive robot control.
APPLICATIONS
UAV planning and guidance, ROV/AUV underwater missions, Multibeam echosounders data collection and control, Intelligent energy harvesting systems, Edge and fog computing appli- cations. Smart devices.
Comments