Use Case Intelligence RSS Search use cases Search
Alibaba Cloud Manufacturing China Translated from Chinese

Midea IoT

IT Operations Automation

Midea IoT depended on an ELK-based device-log pipeline that was becoming hard to scale as connected-device volume and analysis demand rose. The architecture involved too many hops, separate data silos for real-time, offline, and monitoring use cases, and a heavy self-managed Elasticsearch footprint.

Read the original story Explore all use cases
device-log query and bulk data retrieval efficiency improved 50%
self-built ES clusters and log-cleaning services were fully retired

Reported impact

Business outcome

Midea IoT eliminated much of the self-managed ELK overhead and replaced it with a serverless observability substrate that is easier to scale with device growth. Query speed improved, self-built ES clusters and cleaning services were retired, and AI-assisted operations became a native capability of the log platform rather than a separate tool.

  • Cost Reduction
  • Productivity Improvement
  • Risk Reduction

Solution

Midea IoT rebuilt its observability and log-analysis backbone on Alibaba Cloud with SLS as the unified data substrate and STAROps as the AI operations hub. Filebeat was replaced by LoongCollector for direct collection into SLS, Kafka-originated business data is ingested through SLS write processors, and the same data base supports real-time query, subscription consumption, and offline archiving rather than copying logs across several systems. Offline data is compressed into MaxCompute, while OSS and unified data APIs replace much of the old Elasticsearch dependency. On top of the new log architecture, STAROps combines cross-domain observability data with large-language-model reasoning and AI-friendly Skills, CLI tools, and agents so internal monitoring teams can perform natural-language query and root-cause analysis directly against the new substrate. The result is a serverless, lower-ops observability platform that scales with IoT growth.

Data flow

Device logs and business Kafka data are collected into SLS through LoongCollector and SLS write processors, then exposed through APIs, subscriptions, and direct platform reads for monitoring and analysis. Historical data is archived into MaxCompute and OSS, while STAROps reasons over the cross-domain observability data for natural-language analysis and incident diagnosis.

Solution architecture

4 components · 4 layers
  1. Ingest
    • LoongCollector Collects logs directly into SLS in place of Filebeat-based ingestion.
  2. Compute
    • MaxCompute Receives offline compressed archives and supports large-scale analysis on historical log data.
  3. Storage
    • Log Service SLS Acts as the unified ingest, query, consumption, and archival substrate for device logs.
  4. Orchestration
    • STAROps Provides AI-assisted natural-language query and root-cause analysis on top of unified observability data.

Architecture clues

  • Historical archives are pushed into MaxCompute and OSS instead of being left in a hot-search cluster indefinitely.
  • SLS is treated as the single data substrate for real-time query, subscription, and archive flows instead of duplicating logs into multiple systems.
  • STAROps is tightly integrated with Midea's internal monitoring systems so natural-language query and diagnosis work on the same unified data base.

Evidence from the source

After the redesign, Midea reported 50% faster log-query and bulk-retrieval efficiency while enabling natural-language query and root-cause analysis through STAROps. (original: 设备日志查询与大数据拉取时效提升 50%；同时通过 STAROps... 支持自然语言智能查询、根因分析等 AIOps 智能运维能力。)
Midea IoT said its ELK-based device-log pipeline was under pressure as connected-device scale and analysis demand increased. (original: 美的 IoT 联网设备体量大，增长快，业务对设备日志的分析需求有重要依赖... 随着业务规模增长，面临一些技术挑战。)
The company rebuilt the stack with SLS as the unified data base, LoongCollector for direct collection, MaxCompute for archive, and STAROps as the intelligent-operations hub. (original: 以 SLS 为统一数据底座，以 STAROps 为智能运维中枢... Filebeat 替换为 LoongCollector 直采入 SLS... 离线数据投递 MaxCompute 压缩归档。)
Transparency Evidence & methodology 90/100 reference value · 88% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value90/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact15/25
    1 metric describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence88%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability25/25
    12 of 12 key fields link to a named source.
  • Direct evidence15/20
    3 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. Alibaba Cloud

    美的IoT基于阿里云SLS+STAROps重构设备日志链路及智能运维_制造_阿里云客户案例

    First-party
    https://www.aliyun.com/customer-stories/manufacturing-2026-midea
    Source type
    Vendor-published customer story
    Supports
    • Customer
    • Region
    • Company size
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes

Date evidence

  • Published on source page Sep 16, 2026 · Alibaba Cloud · First-party
    Open date source
Not separately sourced Industry. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this
Alibaba Cloud Oct 12, 2026 ECLICKTECH (易点天下) Software / SaaS China ECLICKTECH needed to industrialize AI-driven advertising operations, but media-delivery workflows were still long, manual, and heavily dependent on senior operators for anomaly diagnosis and creative adaptation. As more… AI covers about 50% of core execution work IT Operations AutomationMarketing Personalization AdClaw · AgentCore · AgentLoop Alibaba Cloud Oct 8, 2026 ECLICKTECH (易点天下) Software / SaaS China ECLICKTECH's operations team had to manage a multi-cloud, multi-product environment that supported several lines of business, but its observability data, alerting systems, and AI tooling were fragmented across separate… connects more than 600,000 media outlets globally IT Operations Automation AgentCore · AgentLoop · Cloud Monitor 2.0 Alibaba Cloud Sep 20, 2026 Fintopia (瓴岳科技) Financial Services Global Fintopia had already embedded AI skills across acquisition, risk review, settlement, operations, and internal support, but hundreds of skills were becoming fragmented assets. Teams kept rebuilding similar capabilities,… connected 50+ models Application RefactoringIT Operations Automation AgentLoop · LoongSuite Collector · MCP Plaza