According to the account below, the grand finale of the “2025 Feishu AI Efficiency Pioneer National Competition,” hosted by Feishu, was held in Shanghai on December 3. In this enterprise‑level practical competition, which the organisers position as a pioneering event for “AI + real business impact,” Shanxi Pengfei Group Co., Ltd. (hereinafter referred to as “Pengfei Group”) was reportedly recognised for its self‑developed “AI Full‑Dimension Monitoring System.” The competition name, date, venue and Pengfei's participation should all be confirmed against Feishu's official announcements and award level verified before relying on them. The project, built around a model of “no personnel on site, data transmitted in real time,” offers one possible digital transformation reference for safety production management in the chemical industry.
Hitting the Pain Points:
Defusing the Three “Hidden Volcanoes” in Chemical Workshops
Pengfei Group describes itself as a modern enterprise group spanning coal, coking, hydrogen energy and other industries; readers should check this scope against the company's corporate filings. In the account below, the group treats safety production as the lifeline of the chemical industry. The account then frames traditional chemical workshops as harbouring three major management pain points — a vendor's perspective that buyers should weigh against their own site assessments:
-
High safety risks: Delayed response to gas monitoring in high‑risk areas can easily lead to accidents.
-
Low manual inspection efficiency: Inspecting critical equipment is time‑consuming and prone to omissions.
-
High technical barriers: Incompatible equipment protocols make it difficult for new employees to get started.
The account attributes the remark “In a chemical workshop, every extra minute you stay means more danger” to Shang Yangjian, described as an instrumentation inspection engineer at Pengfei Group and the competition participant; both his role and the quotation should be verified against the original source. As told in the account, he led his team in developing the “AI Full‑Dimension Monitoring System” using Feishu and AI technology, aiming to build a safety defence line through technical means.
Technological Breakthrough:
AI Makes 2,111 Monitoring Points “Speak”
Pengfei Group’s “AI Full‑Dimension Monitoring System” is not a simple stack of equipment, but a deep reshaping of the production process. The system is designed to integrate multi‑source heterogeneous data from PLC, DCS, GDS, UPS, and other systems, helping to break down the “information silos” across the plant’s monitoring points.
Through deep integration of Feishu multidimensional tables and AI capabilities, the system achieves three major breakthroughs:
-
Real‑time brain: Gas concentrations and equipment status are unified and visualised. Decision‑makers can grasp the entire plant’s “pulse” at any time.
-
Rapid diagnostics: When data anomalies occur, the AI is designed to complete self‑diagnosis quickly and automatically generate standardised resolution plans.
-
Ultra‑fast alerts: Alerts are routed to responsible persons through a dual mechanism of “@message + phone call” via the Feishu bot, with the aim of materially shortening emergency response times. Whether response improves as claimed is worth verifying with on‑site incident logs during a trial.
Effectiveness Validation:
Reported Efficiency and Safety Gains
During the evaluation at the Feishu AI Efficiency Pioneer Competition, Pengfei Group presented real‑world project data, which the team reports was well received by the judges. The launch of this system has brought significant dual benefits – both safety and economic – to the Group:
-
Efficiency revolution: Equipment inspection time has been substantially compressed, delivering significant working‑hour savings for the Group.
-
Safety transformation: Maintenance personnel spend less time physically entering high‑risk areas, reducing personal exposure and helping to lower equipment out‑of‑control risk.
-
Cost reduction & efficiency improvement: The team reports meaningful cumulative cost savings, partly from reduced downtime. Treat the specific yuan figures as vendor claims until supported by audited financials or customer references.
Industry Response:
From “Experience‑Driven” to “Data‑Driven”
Pengfei Group’s practice is presented by the team as a regional benchmark for deeper integration of AI with industrial equipment, and as an example of the chemical industry’s broader shift from “experience‑driven” to “data‑driven” management. The team also reports that the case has received earlier recognition, including industry and provincial innovation awards and showcase appearances. These awards should be verified against official award records before republication.
Feishu leadership commented during the finals that, in the AI era, the protagonist is not technology – it is people. Pengfei Group’s entry is offered as an illustration of this philosophy: frontline business personnel, using AI tools, aim to turn on‑site experience into digital assets and reshape their business processes. Going forward, Pengfei Group says it intends to deepen AI use in production management and explore whether the “AI Full‑Dimension Monitoring System” can become a replicable industry‑level solution, with the aim of improving safety and efficiency for other traditional manufacturers.
Thinking it through before you copy an award-winning idea
A CEO or operations manager in Hong Kong reading about this competition entry should separate the idea from the execution. The concept — pulling readings from plant equipment into one shared view so people spend less time walking the floor — is attractive in any hazardous-workplace setting, whether a chemical plant, a data centre utility room or a food-processing line. The harder question is whether your own equipment, staffing and escalation habits can support something similar.
Start by mapping the current workflow with your plant engineering team: which machines are monitored today, how readings are recorded, who responds when something drifts out of range, and where delays usually occur. Then name one owner for the trial and load fictional readings into a sandbox environment to see how alerts would actually reach the duty team. Ask each vendor for a written answer on how the system behaves with the equipment protocols you already run — and note anything that remains unconfirmed rather than assuming it will work out.
Unanswered questions about where information is stored and how it moves between Hong Kong and mainland sites should go to your privacy, security or contract owner before any rollout. If your teams operate on both sides of the boundary, our piece on cross-border collaboration and data governance for HK and Macau teams is a useful next read.