The building blocks of a smart city that seeks to create a people-oriented management system based on advanced technologies, including the Internet of Things, cloud computing and artiﬁcial intelligence. Our solutions for smart communities and industry, and are based on compute vision, gesture recognition and machine learning algorithms, integrate segmented scenarios to provide a complete solutions ranging from target recognition, behavior detection, area monitoring and smart park operation,to make people’s work and lives more convenient, comfortable and eﬃcient.
Car Park Management
—— Vision Technology ——
Software-and-hardware integrated products
—— Auto Machine Learning Platform ——
Landing AI solutions into different scenarios
A Community in Beijing
City Brain System:
Lawbreaking Scenario Detection Sub-system：Detection Framework, Control Monitor, Video Stream Access, Detection Result Storage, Adlet Detection, Tourists, Merchants and Crowd Detection, Log Management, Starting Up
Control and Management Sub-system：Single-channel Video Real-time Display, Detection Start/Stop Control, Detection Result Alarm, Detection Result Display, Detection Result Query, Tourists, Merchants and Crowd Detection Parameter Setting
—— PUE Management ——
Leverages machine learning and deep learning algorithms to accurately forecast PUE, power and other environments, to provide precise PUE awareness for operational staﬀ.
Uses machine learning and decision optimization algorithms to identify the parameter and to identify the best portfolio of operating equip -ment to reduce PUE expenditures.
Acts as a dashboard for operative staﬀ to monitor day-to-day equipment operative, KPI changes, malfunction and other important variable changes in real time.
Provides an optimized strategy for operation parameters of equipment portfolios, feedback fro, model learning, which supports self-learning and upgrading within a closed-loop of control.
Fixes and upgrades and the existing BA system, adds or upgrade existing sensors and communication devices to provide adequate data support for algorithms.
- Real Estate
- Data Center
- Precise PUE forecast; error within 10e-4
- Fine-grained control interval (<5 minutes)
- Automatic, unmanned, closed-loop equipment control
- Real-time response to IT load and changes in external emperature and humidity
- Accurate sub-system modeling and solutions for global optimal control point
- MPC control