Skip to main content
Bivrost 彼络
Bivrost Cloud · 彼络云

Machine data and maintenance workflows,
managed in one platform

Data from the gateway becomes monitoring, analytics, asset management and program transfer in Bivrost Cloud; public cloud, private cloud and on-premises run the same code. Data is exposed to upstream applications through standard interfaces and MCP forwarding.

Monitoring & dashboardsAnalytics & export5 asset-management modulesProgram transfer3 deployment options
Monitor

Live state of machines, cells and tools, plus summary dashboards

Analyze

OEE, output, alarm and combined analytics with export and charts

Manage

Registry, maintenance workflows, spare parts, console, configuration

Monitoring & dashboards

From one machine to the whole floor,on one screen

Standardized points from the gateway flow straight into the platform — no second modelling step.

Machine & cell list

The home page lists every machine by cell with state at a glance; open any machine for detail.

Machine detail

State, current program, tool number, load, feed, speed, output and utilization, refreshed live.

Cell detail

State distribution and output totals across the machines in a cell.

Tool detail

Tool numbers and tool life, paired with offset read-write.

Summary dashboard

Floor-level utilization, output and alarm totals for wall screens and shift meetings.

Floor map

Machine positions and states on the shop-floor layout.

Analytics

OEE, output and alarms,sliced by time and cell

Results export; charts and tables convert into each other.

OEE
Availability, performance and quality broken out by day, week, month and cell
Output
Output and cycle time by machine, cell and shift
Alarms
Alarm count, duration and type rankings to find the machines that keep failing
Combined
State-time distribution and run trends, with export and chart conversion
Asset management

Registry, maintenance, spares, console, configuration —five modules, one record

Asset-management data and live machine data share one platform; repair, inspection and maintenance records attach directly to the machine history.

01

Registry & lifecycle

Assets, lines and stations on record, one code per machine; history, stock-takes, transfers and issues all logged.

02

Maintenance workflows

Scan-to-report repairs, inspections, three-level preventive maintenance and QC sign-off, every ticket traceable.

03

Spare parts & consumables

Parts registry, receipts, issues and stock-takes; low-stock alerts.

04

Console

Live monitoring of gateways, cells and machines with alarm history.

05

Configuration

Notifications, forms, lines, product models, processes, users, shifts and handovers.

Program transfer

Program files and directories, read and written from the cloud

Read, upload, download, delete, back up and bulk-send machine programs through the gateway. Shops that need versioning, approval and controlled deployment use DNC Pro.

DNC Pro product page
Open

What runs on top of the platformis your decision

Bivrost Cloud is not where the data stops. Your own dashboards, a developer’s agent or a general-purpose AI assistant all use the same interfaces.

Standard interfaces
Platform data exposed over HTTP, MQTT and other standard interfaces
MCP forwarding
The platform-side MCP server forwards requests to the chosen gateway: one connection for the AI client, many gateways reachable
One auth model
Keys plus an authorized-API allow-list, granted per application, per endpoint
Developers page
AI application

Factory AI:ask the platform’s data in plain language

Bivrost’s own AI application, built on the same platform and shipped with Bivrost Cloud. Every capability it uses is exposed through the same gateway interfaces to anyone else.

tb.bivrost.cn / factory-ai Open the workbench
Factory AI workbench: project sidebar, conversation area and the instructions / files / skills / scheduled tasks panel
Weekly review conversation: alarm counts, total downtime and daily efficiency per cell, with the bottleneck called out
01

Weekly bottleneck review

Finds the week’s bottleneck cell from alarm counts and downtime, quantifies the gap to target; can run as a weekly scheduled task.

Template picker: analytics, production management and cycle-time improvement system templates
02

OEE, output and cycle-time analysis

Start a project from a template — analytics, daily production brief, cycle-time improvement — with skills, instructions and scheduled tasks preset.

Repair knowledge Q&A: past work orders searched, similar faults listed with fixes and parts
03

Repair history, answered on the spot

Past faults on the same model, how they were fixed and which parts went in — retrieved straight from repair work orders.

Measured at a customer · Jiehe Motor

90 machines, two sites, one record across three roles

60 machines in Shenzhen and 30 in Jiangxi on 3 gateways, over 9,000 points in the platform. 55 accounts across management, maintenance and operator roles; repair requests, dispatch and sign-off close inside the platform.

90
machines on 3 gateways
9,000+
collected points
55
platform accounts across three roles
1,000
repair work orders per month (Shenzhen, 60 machines; excludes inspection exceptions)
48
inspection tickets auto-dispatched per shift
−30%
ticket-creation time (equipment manager’s estimate, creation step only)
Deployment

Public cloud, private cloud, on-premises —one codebase

Choose the deployment by your data-governance requirements; no feature is cut.

Public cloud

Ready to use; a gateway with internet access is connected.

Private cloud

Deployed in the cloud environment you specify; data stays in your account.

On-premises

Runs on a server inside the plant, isolated from the internet.

Book a 30-minute demo

With your own machines: see how data enters the platform and how it is managed.

Submitting opens your mail client with the message pre-filled; you can also write to contact@bivrost.cn directly.