新闻中心

位置 > 首页 > 新闻中心 > News

Teamsun Smart Park has successfully connected to DeepSeek, and the park’s intelligent experience has been “dimensionally enhanced”!

时间:2025-02-10 19:03 浏览量:2028

On February 9, the baseline version of Teamsun Smart Park has been fully integrated with DeepSeek open source large model technology, creating a new generation of AI-driven smart park solutions through independent development and scenario optimization. This move marks an important breakthrough for Teamsun in the field of integration of open source technology ecology and industry scenarios, aiming to improve park management efficiency, service model innovation and sustainable development with AI capabilities at lower costs and higher flexibility.


image
01
technological independence



Open source large models drive intelligent upgrading of smart parks

Smart Park Baseline Technology Architecture Diagram

In the context of the accelerated open source trend of AI technology, Teamsun chose to use the DeepSeek open source large model as the technical base based on its in-depth understanding of smart park scenarios to build exclusive scenario-based AI capabilities through independent training and vertical domain optimization. Company technical teamAround the park managementAddressing pain points, combined with Huawei's digital platform, the directional enhancement of the large model from general capabilities to industrial knowledge has been completed, and the value reconstruction of three core scenarios has been achieved:



Cross-system semantic collaboration breaks down data silos



Relying on the multi-modal understanding and reasoning capabilities of DeepSeek's large model, combined with the connection services of Huawei's digital platform, Teamsun Smart Park Platform achieves unified semantic analysis of subsystems such as energy, security, and equipment operation and maintenance. For example, administrators can use natural language instructions to "check the abnormal energy consumption area on the third floor of Building A this week" to trigger cross-system linkage. The platform automatically correlates data from air conditioners, lighting, elevators and other equipment to generate visual analysis reports, increasing decision-making response efficiency by 50%.




Dynamic resource scheduling to achieve green and low carbon



In traditional campuses, energy management and equipment operation and maintenance are highly dependent on manual experience. Teamsun combines the data services of Huawei's digital platform and integrates the timing prediction and optimization algorithms of the DeepSeek large model. The system can analyze weather, people flow, equipment status and other data in real time, and dynamically adjust the air conditioning temperature, lighting brightness and charging pile power. A pilot project in a manufacturing park showed that after accessing the AI ​​model, monthly comprehensive energy consumption was reduced by 18%, and annual carbon emissions were reduced by more than 1,500 tons.



Natural interactive services to improve user experience



Operation center system diagram based on AI large model

In response to the needs of park employees and visitors, Teamsun has developed applications such as intelligent voice assistants, AR navigation, and personalized service recommendations based on open source models.Users can directly complete conference room reservations, visitor registration, repair applications and other operations through dialogue, and the service interaction time is shortened by 60%.At the same time, the system supports Chinese-English bilingual and dialect recognition, covering more than 90% of daily demand scenarios.


02
Eco-innovation



 Open source technology + industry Know-How two-wheel drive

Zhang Ying, Vice President of Teamsun, said: "Open source large models provide the industry with inclusive AI infrastructure, but the real value lies in how to deeply integrate it with vertical scenarios. Teamsun is based on the DeepSeek open source framework and Huawei digital platform,Injecting more than 200,000 pieces of park management knowledge data and developing lightweight deployment tools enable AI capabilities to quickly adapt to the personalized needs of parks of different sizes. "


03
Industry value



Open source ecology helps make smart parks inclusive

According to IDC's forecast, China's smart park software market will reach 48 billion yuan in 2025. However, small and medium-sized parks generally face challenges such as limited budgets and a lack of technical talents. Teamsun’s upgrade this time uses open source technology to reduce AI deployment costs, and at the same time provides standardized solutions and flexible functional module combinations, allowing park customers to “call on demand and gradually upgrade” according to needs. At the same time, Huashengtian has ensured technological independence and built a complete data security system based on the open source model. All park operation data are processed through locally deployed privacy computing modules, ensuring that the original data does not leave the domain. In addition, Teamsun has also established an AI decision-making traceability mechanism. Key instructions require secondary confirmation by the administrator, avoiding the risk of large model illusions.

In the future, Teamsun will uphold the spirit of open and win-win cooperation, join hands with more ecological partners, continue to deepen the use of the open source ecosystem, increase the in-depth integration and adaptation of smart parks and DeepSeek, provide more diverse and flexible application options for the integration of enterprise AI technology and specific scenarios, fully meet the needs of users, and promote the in-depth development of my country's digital and intelligent transformation.


版权所有©北京华胜天成科技股份有限公司 京ICP备14026093号-1
在线客服