How Zhejiang Mobile’s AIOps Achieved National‑Level Excellence in Fault Management

What is AIOps?

Intelligent Operations (AIOps) applies artificial‑intelligence techniques such as machine learning and data science to IT‑operations problems, enhancing and partially replacing core IT‑ops functions. Gartner describes AIOps as a loosely‑coupled, scalable approach that extracts and analyzes ever‑growing volumes, varieties and velocities of IT data.

Forum and Assessment Results

On 24 December, the China Academy of Information and Communications Technology (CAICT) launched the 2021 GOLF+ IT New Governance Leadership Forum. The event announced the first official evaluation results of AIOps systems and tools. Zhejiang Mobile’s fault‑center project passed the “Cloud Computing Intelligent Operations (AIOps) Capability Maturity Model – Part 2: System and Tool Technical Requirements” assessment, with the fault‑prediction module receiving a full‑level rating, indicating a domestic leading level.

Interview Highlights

In an interview, Zhu Shijie, Deputy Director of Zhejiang Mobile’s Network Management Center, explained that Zhejiang Mobile has built a “135” intelligent fault‑management framework: a closed‑loop fault‑management system (1), three alarm‑knowledge rule libraries (3), and five core capabilities (5) covering open APIs, configurable orchestration, visual alarm‑quality monitoring, large‑scale storm handling, and intelligent operations. He said the assessment both validates their work and provides a roadmap for further AIOps improvement, improving team skills and operational mindset.

He also outlined three future plans: expand AIOps applications (e.g., smart energy management for data centers and 5G stations, voice‑quality analytics, network‑infrastructure QA), build a national‑level AIOps platform to enable data sharing, and integrate AIOps with the carrier‑wide network autonomous‑driving initiative.

Future Directions

Zhu emphasized that the rapid rollout of 5G and its convergence with industry creates unprecedented operational complexity, demanding AI, machine‑learning and big‑data solutions. He recommends industry‑wide unified goals, a shift for ops staff toward DOICT (Data‑center‑Oriented IT‑Operations) skills, and ecosystem‑driven standardisation to accelerate technology adoption.

AIOps Capability Maturity Model

The “Intelligent Operations (AIOps) Capability Maturity Model” was jointly authored by CAICT, the Cloud Computing Open‑Source Industry Alliance, the Efficient Operations Community, BATJ and leading telecom and financial enterprises. It is the first domestic and international standard for intelligent operations, approved by ITU‑T SG13. The pilot assessment currently opens four quality modules: anomaly detection, fault prediction, alarm convergence, and root‑cause analysis.