[Shanghai, China, September 19, 2019] During HUAWEI CONNECT 2019, Huawei officially launched the Network AI Engine (NAIE) for operators, such as China Telecom, enterprise partners, universities, and other network AI developers. This marks another milestone for Huawei in using AI to enable autonomous driving networks (ADNs).
Han Yufa, General Manager of Huawei NAIE Product Dept, introduces the NAIE
Huawei NAIE provides the data lake service for data governance and the model training service for model generation on the public cloud. These services solve the most complex tasks during AI application development. Han Yufa, General Manager of Huawei NAIE Product Dept, announced the commercial release of the NAIE at the news conference. This includes the following four cloud services that are required for network AI application development: data asset management, model training, model generation, and communication model.
Data Asset Management Service
This service provides data processing tools for data collection, integration, modeling, analysis, labeling, and other purposes, and it also provides data governance templates as cloud services for developers to improve data governance efficiency.
Model Training Service
This service provides an integrated development environment (IDE) for data processing, feature extraction, model training and verification, and other functions. This service integrates the 30 years of knowledge and experience that Huawei has accumulated in the network domain. SDKs designed for algorithm commissioning and feature services and processing significantly shorten the model design and exploration period.
Model Generation Service
This service further simplifies the model development process. Developers can simply offer training data in a preset typical scenario model to rapidly train, verify, and generate an AI model. This simplifies telecom AI model development and shortens the development period.
Communication Model Service
This service relies on a cloud-based inference framework. Users can rapidly perform inference by merely providing inference data. The inference result can be used for service application development. This service is applicable to general models, such as disk fault detection. Since most disk fault symptoms and characteristics are the same, users can provide SMART data to acquire disk health check results online. Similar scenarios include KPI anomaly detection and ECA anomaly detection.
Han Yufa, General Manager of Huawei NAIE Product Dept, and Dong Zekun, Director of Huawei-China Collaboration on Telecom Network AI Innovation, jointly launch NAIE
Based on Huawei NAIE, China Telecom has developed AI application development capabilities in the e-Surfing Cloud. This enables the company to provide a diverse range of network AI cloud services and use AI technologies in different telecom services. In addition, China Telecom has deployed AI applications to improve O&M efficiency and energy efficiency in many regions, such as Guangdong. Dong Zekun, Huawei-China Collaboration on Telecom Network AI Innovation, reviewed China Telecom's network AI project, looked into the future of AI applications in base station energy, core network KPI anomaly detection, as well as other fields, and expressed his confidence in the benefits offered by AI applications. He hopes more industry partners and developers will create innovative network AI applications.
Liu Kai, Chief Architect of Huawei NAIE Product Dept, demonstrates the NAIE training platform
How can the optimal hyperparameter settings be identified with regard to limited computing power, cost, and time? Liu Kai, Chief Architect of Huawei NAIE Product Dept, shared theories and practices on common AI development issues. Liu Kai first introduced the common hyperparameter search policy from an algorithm perspective, and detailed the Bayesian optimization principle.
The automatic hyperparameter optimization service of the iMaster NAIE supports random search, grid search, SMAC, and Bayes optimization. This service fully leverages the parallel training capability of cloud computing and simplifies parameter adjustment for AI developers. Liu Kai also introduced the iMaster NAIE training platform and how to invoke APIs, demonstrated how to implement visualized feature engineering on the NAIE training platform, and used the automatic hyperparameter optimization service to optimize the xgboost model.
First prize winner of the NAIE Track in Huawei Developer Challenge Jiang Ting shares her experience of using the NAIE training platform
Jiang Ting, the first prize winner of the NAIE Track in Huawei Developer Challenge, shared her experience of using the NAIE platform. Her AI team used the NAIE platform to develop an AI model for reducing DC energy consumption. This model fits and predicts the power consumed by DC cooling systems to lower the power consumption of entire cooling systems and save energy. Jiang Ting commented: " The NAIE platform is easy to use, even for developers who have never used it before. The integrated development environment (IDE) makes the data exploration, feature engineering, and algorithm commissioning processes clear. The platform integrates abundant network AI knowledge and experience, making AI development simple and efficient. "
Han Yufa, President of Huawei NAIE Product Dept, said, "Huawei NAIE is designed to facilitate the usage of AI, improve AI development efficiency for network services, and simplify network AI development for operators, enterprise partners, universities, and other developers. Huawei is dedicated to jointly improving the AI industry environment and deploying an ADN with industry partners."