A digital twin
of production:
from monitoring to forecasting
Procesia Industrial Cloud is a cloud platform that builds a digital twin of your shop floor. It sees every event, detects deviations, and forecasts downtime risks before they happen. It connects to your existing MES (SIMATIC IT / Opcenter) with no changes on the plant side.
Procesia Industrial Cloud
A digital twin of production: real-time monitoring of shops and equipment, MES events, ML downtime forecasting and component demand forecasting with purchase requisitions exported to 1C. Connects to SIMATIC IT / Opcenter with no changes on the plant side.
PLATFORM FEATURES ↓ FLEETProcesia Fleet — vehicle fleet monitoring
GPS/GLONASS vehicle positions on a live map, track history, and bulk-cargo unloading control: a cargo-level sensor in the body shows the tonnage of every trip in real time.
MORE ABOUT FLEET MONITORING →Everything you need for transparent production
Connect to your existing MES with no changes, web dashboards, a downtime log with acknowledgement, Grafana analytics, and ML downtime forecasting based on acknowledgement history.
Equipment monitoring
A shop-floor layout with real-time status indicators. Click any element to see its details and events.
Downtime log
A CRUD log with acknowledgement. The operator selects shift, cause and comment. XLSX export by period.
Grafana analytics
Ready-made dashboards: downtime dynamics by period, top causes, OEE and line activity, event stream.
Multi-plant architecture
Each plant has its own agent and access token. Data isolation — an owner sees only their own plants and shops.
QR codes for tablets
A QR code on a shop-floor tablet → instant access with no login. Adapted for mobile screens.
Agent auto-update
Centralized version management. Agents update themselves, with a pinned version per plant.
Predictive AI analytics
Prediction of downtime and anomalies before they occur, based on operators' downtime-acknowledgement history. Models are trained separately for each plant.
Component demand forecasting
The system analyzes write-off and component-consumption history, computes consumption rate, safety stock and the reorder point — and signals in advance what to purchase and when. An ML model refines the forecast using consumption trends and seasonality, and the generated requisitions are exported to 1C automatically.
Consumption rate
Calculates the actual consumption of each component from write-off history: per shift, week, month. Accounts for seasonality and peaks.
Reorder point and safety stock
For each item — the threshold at which to place an order, factoring in lead time. A signal before the stock hits zero.
Shortage forecast
How many days the current stock will last at the present consumption rate. Early warning of a line-stoppage risk.
Requisitions to 1C
Automatic supplier-order generation for items past the reorder point. Exported to 1C with no manual entry.
Procurement dashboard
An overview of all components: stock, consumption rate, days to shortage, requisition status. Filters by warehouse and category.
ML forecast refinement
The model accounts for consumption trends and seasonality more accurately than plain statistics and continuously retrains on new consumption data.
Live previews of the running system
These previews open straight from the production server — what you see here and what your operator sees are always the same.
Delivered automation and optimization projects
MES connection, reconfiguration of existing production setups, integrations with weighing and laboratory systems. Below are a few examples of completed work.
Connecting a new plant to the MES
SIMATIC IT integration with cloud monitoring, agent setup, migration of equipment and operator reference data. Timeline — 4 weeks from start to the first dashboard in production.
Plant configuration optimization
Audit of existing recipes, reconfiguration of routes and production nodes, reduced changeover time between batches. Before/after measured on real data via the built dashboard.
MES reconfiguration for production expansion
Adding a new line to a live MES, configuring reference data, integrating new sensors and operators. No stoppage of the running shop, staged migration.
From sensor to dashboard — in seconds
A Windows agent on the plant side pulls data from MSSQL and sends it to the cloud over an encrypted channel. No changes to the MES, no VPN, no open inbound ports.
SIMATIC IT / Opcenter
HTTPS · JWT · API key
Docker Compose
dashboards · filters
Downtime forecast
downtime history
7 steps from audit to forecast
A standard implementation cycle is about 4–6 weeks: from an MES audit to working dashboards, AI analytics and procurement forecasting in production.
Audit
MES analysis, SQL Server version, network availability, the list of tables to poll.
Agent
Windows agent installation, table polling setup, NSSM service, auto-update.
Cloud
Deploy API + DB + Grafana on a VPS or your cloud. Docker Compose, SSL.
Dashboards
Shop-layout setup, QR tokens for tablets, the operator and cause reference data.
Pilot
Testing on real data, training operators to acknowledge downtime.
Scale
Connecting additional plants and shops, data isolation, access control.
AI analytics
Downtime and component-demand forecasting, purchase requisitions exported to 1C.
MES ↔ 1C — production and accounting work as one
Beyond cloud monitoring, our team delivers MES-to-1C integrations as well as AutomationX customizations tailored to a specific production site.
- Manual data transfer between the MES and 1C
- Entry errors, accounting delays
- Production and finance out of sync
- Decisions made with a lag
- Direct MES ↔ 1C exchange with no operator
- Automatic transfer of output and consumption
- Reconciliation of item reference data
- Exchange log and error control
- SIMATIC IT (axpdb) as the data source
- 1C UPP / ERP as the receiver
- Exchange bus over XML / web services
- Scheduled jobs and monitoring
- Human error eliminated
- Faster response
- Faster decision-making
- Production and accounting work as one
- SCADA/MES tuning for a specific plant
- New operator screens, reports and logs
- Refinement of alarm and signaling logic
- Integrations with weighing systems, ERP, the lab
- Incident resolution and bottleneck removal
- Regular performance audits
- Component updates and migrations
- Training for operators and engineers
Customization per client — no queue, no "that's not possible"
Large platforms deliver customizations over months of bureaucracy, or simply answer "that's not supported." Procesia is independently developed with full control over the stack: we build the feature, report or integration you need around your process, instead of forcing your process into a box.
Changes in days, not months
Our own stack and direct contact with the developer. A new feature or fix for your specifics — without long approvals and vendor ticket systems.
Reports and exports for you
Forms, fields and formats the way your production needs them. PDF and Excel export tailored to your regulations and auditors.
Integration with your accounting
1C, ERP, weighing and laboratory systems, internal databases. The exchange is configured to your specific reference data and exchange rules.
Sensor calibration per site
Level, flow, weight, telemetry — calibrated to your equipment and actual vessels. Support for industrial sensors and protocols.
Your data on your perimeter
Self-hosted on a dedicated server. Data does not go to a third-party vendor's shared cloud — critical for industry and data-residency compliance.
Not just a login — a process
Beyond access to the system — we help build the monitoring and procurement process itself: what to measure, how to respond, which thresholds to set.
Frequently asked questions
The most common questions about connection, security, supported MES systems, timelines and ML forecasting.
Do we need to modify our existing MES to connect Procesia?
No. The Windows agent runs read-only over ODBC: it periodically polls MSSQL tables by a datetime field. No changes to SIMATIC IT / Opcenter are required, no CDC or triggers needed. Zero load on the live system.
Does the plant side need a VPN or open inbound ports?
No. The Windows agent only initiates outbound HTTPS connections (port 443). No inbound ports need to be opened on the plant firewall. Authorization via API key, TLS 1.3 encryption.
Which MES systems are supported?
Initially optimized for Siemens SIMATIC IT and Opcenter. It also connects to any MES storing data in MSSQL: AutomationX, custom systems with an MSSQL backend. Table polling is configured to a specific MES schema without any coding.
How long does implementation take?
The standard cycle is 4–6 weeks from the first technical call to a working dashboard in production. It includes an MSSQL audit, cloud deployment, agent setup, reference-data migration, and operator training. See the case studies for details.
How does ML downtime forecasting work?
The forecast is trained on operators' downtime-acknowledgement history and predicts downtime and anomalies before they occur. Grafana analytics run in parallel — trends, OEE, top causes. The model is trained separately for each plant (per-plant).
Can the platform be used for multiple plants?
Yes. The architecture is multi-plant by design: each plant connects via its own agent with its own API key. Data is isolated — an owner sees only their own plants. ML forecasts are trained per plant (per-plant models).
How does component demand forecasting work — is it AI?
The system calculates consumption rate from write-off history, safety stock and the reorder point, while an ML model accounts for consumption trends and seasonality. The platform signals in advance what to purchase and when, and exports purchase requisitions to 1C.
Can the system be customized to our process?
Yes — this is the key difference. Procesia is independently developed with full control over the code, so reports, exports, 1C/ERP integrations and sensor calibration are tailored to your specifics, usually in days rather than months. The scope is agreed during the audit stage.
Ready to discuss deployment at your production site
The first 30 days of the pilot are free. Reach out any way you like — we'll discuss your MES infrastructure and put together an implementation plan.