PlantPulse
This sheet summarizes value, pricing and use cases. From here you can move to the dedicated areas or to the client portal for the next action.
PlantPulse measures energy directly on the field and transforms it into targeted maintenance and demonstrable savings. Propri hardware nodes or your existing PLCs via Modbus, MQTT and OPC-UA: from the sensor to the signed report for audit.
What it solves
- Electric consumption is invisible until the bill arrives: no one knows which machine is wasting energy, nor since when
- Maintenance is either calendar-based or reactive: intervention happens too early, wasting resources, or too late, resulting in unplanned downtime.
- The declared energy savings cannot be demonstrated to a customer or an auditor: baselines, climate normalization, and uncontested evidence are missing
- The data is stored in incompatible silos: PLCs, counters, ERPs, and spreadsheets do not communicate with each other, and connecting them requires opening incoming ports on the OT network
What it does
PlantPulse closes the loop from field measurement to maintenance decision and savings proof. In the field, nodes based on Raspberry Pi with ADC MCP3008 read voltage and current, calculate power and cos phi, and send batch measurements to the central server with buffer and retry; alternatively, an on-site gateway queries installed PLCs and counters via Modbus TCP and OPC-UA, or the server subscribes to an existing MQTT broker. The server persists, aggregates, and evaluates every measurement in real-time: exceeding a threshold triggers an alarm, deviating from the expected electrical signature generates a predictive work order, falling below the baseline energy level becomes certifiable savings. All traffic is outbound and authenticated with per-device key: no need to open incoming ports on the customer's firewall.
| Module / component | Role |
|---|---|
| Edge node (Raspberry Pi + ADC MCP3008) | Reads V/I via SPI, calculates power and cos phi, buffers and sends HTTPS batch |
| On-site protocol gateway | Polls Modbus TCP and OPC-UA to existing PLC/meters with config downloaded from server and hot-reload |
| MQTT Bridge (server side) | Subscribes to customer brokers; recomposes V/I/P and feeds measurements and tags |
| Application Server | ASP.NET Core .NET 10: ingest, domain, alarms, analytics, real-time SignalR |
| Tree-based asset registry | Site, building, installation, machine, component, with energy data, MTBF/MTTR, OEE, depreciation, cost center |
| Maintenance | Work orders, calendar and kWh/hour counter plans, checklists, spare parts, photo and signature attachments |
| Predictive health | Electrical signature deviation and remaining useful life, with automatic work order generation |
| Alerts and anomalies | Sensor and tag thresholds with cooldown, z-score detection, escalation and notifications |
| Savings and M&V | Declared baseline and regression on drivers, EnPI, EPC contracts with reporting |
| Compliance schedule | Recurring document checks (grounding, fire safety, F-Gas) that generate inspection orders |
| Multi-vector ESG | Electricity, gas, water, district heating, diesel, LPG, with kWh factors and kgCO2e |
| Integration | Power BI/Excel OData feed, signed HMAC webhooks, ERP import/export, master data export |
| Web dashboard | Angular 22 standalone tiles, 43 functional areas, separate customer portal |
Key Features
- Real-time monitoring of voltage, current, power and cos phi, with SignalR consumption aggregation and fleet status monitoring.
- Three connection paths without opening the incoming firewall: proprietary edge agent, Modbus/OPC-UA gateway on existing automation, or direct push HTTPS from SCADA.
- Auto-registration of plants and sensors upon first transmission, with web-enrichable directory.
- Threshold alarms (power, voltage, current, cos phi) and generic non-electrical tag alarms, with cooldown, handling and automatic conversion to work order.
- Predictive maintenance driven by energy: deviation in the electrical signature triggers a work order before failure.
- Maintenance plans based on calendar and usage counter (kWh or hours), with standardized spare parts and inventory.
- Measurement and verification of savings according to IPMVP Option C: regression baseline over time, production and degree-days, with R2 and evaluation days.
- Automatic climate normalization: HDD/CDD degree-days are downloaded from the weather service based on baseline latitude and longitude.
- Sealed M&V reports for audits: canonical JSON, SHA-256 hash for integrity and HMAC signature for authenticity, in immutable history and re-verifiable.
- EPC contracts with guaranteed savings or shared savings, automatic accounting, penalty calculation and ESCO/client split.
- EnPI according to ISO 50001: kWh per unit produced against target, with percentage deviation.
- Cross-tenant ESCO portfolio and customer portal with scheduled monthly PDF reports.
- Multi-vector accounting and ESG with emissions in kgCO2e and water consumption.
- Native multi-tenancy with JWT, rotating refresh tokens, roles and audit trail.
- OTA firmware update of nodes: catalog with SHA-256, channels and sensor groups.
- Optional OpenTelemetry observability via OTLP and Prometheus, configurable retention and hypertable TimescaleDB.
Architecture
Field nodes are .NET agents on Raspberry Pi that read the MCP3008 ADC over SPI bus, sample V and I, derive power and cos phi, and batch forward HTTPS to the ingest endpoint, authenticating with a device-key and locally buffering with retry when the network drops. In parallel, a gateway worker lives on the industrial LAN, downloads its configuration from the server with hot-reload, polls PLCs and counters via Modbus TCP and OPC-UA, applying scales, offsets, and datatypes to map each register or NodeId to an electrical field or generic tag; devices publishing over MQTT are consumed by a dedicated service on the server side that subscribes to customer brokers. The data contract between field and server lives in a single shared assembly so message format is guaranteed at compile-time on both sides. The server is ASP.NET Core .NET 10 minimal API with EF Core and Npgsql on PostgreSQL, multi-tenant isolation via query filter, SignalR for real-time push, OData for BI feed, and signed HMAC-SHA256 webhooks with outbox and retry. The front-end is Angular 22 standalone with signals served by nginx. Deployment goes from Docker Compose (TimescaleDB, server, web) to Windows service via Application Hub; edge nodes install as systemd units from a self-contained linux-arm64 package or as arm64 Docker image.
Requirements
- Field Nodes: Raspberry Pi with Raspberry Pi OS Lite 64-bit, ADC MCP3008 on enabled SPI bus, voltage and current sensors; systemd service from linux-arm64 package or arm64 Docker image.
- Protocol Gateway: a machine on the industrial LAN with access to Modbus TCP or OPC-UA PLC/meters and outbound HTTPS connectivity.
- Connectivity: only outbound HTTPS traffic from field to server, authenticated with device-key for device; no incoming ports to open.
- Server: .NET 10 runtime (ASP.NET Core), or Docker and Docker Compose, or Windows as a service via Application Hub.
- Database: PostgreSQL, with optional TimescaleDB for hypertables, compression and retention of time series data.
- Front-end: built with Node.js 24 LTS, served by nginx; at runtime just a modern browser is needed.
- Optional: client MQTT broker, SMTP for notifications, OTLP or Prometheus collector for observability.
Compliance & regulations
Sistemi di gestione dell'energia
Calcolo degli EnPI (indicatori di prestazione energetica) come kWh per unita' prodotta per asset sul periodo, confrontati con il target definito sulla baseline e con scostamento percentuale. Il prodotto fornisce gli indicatori a supporto del sistema di gestione: non costituisce di per se' una certificazione.
International Performance Measurement and Verification Protocol
Baseline a regressione sui driver secondo l'Opzione C (ore, produzione, gradi-giorno) stimata ai minimi quadrati, con risparmio 'a parita' di condizioni', giorni effettivamente valutati e R2. Gradi-giorno HDD/CDD normalizzati automaticamente da dati meteo.
Reports & dossiers
- Report M&V firmato per auditJSON canonico + SHA-256 + firma HMAC
Sigilla il risparmio di un periodo in uno storico immutabile: l'hash e' verificabile da chiunque, la firma attesta l'autenticita'. La prova da mettere sul tavolo di un auditor o di un cliente EPC.
- Report ESCO mensile per clientePDF
Risparmi M&V, emissioni e manutenzione del periodo per singolo cliente, con PDF conservato e hash per l'audit. Su richiesta o pianificato, inviato via email e scaricabile dal portale cliente.
- Rendiconto contratto EPCVista di rendiconto stampabile
Confronta il risparmio reale con la garanzia pro-quota: percentuale di raggiungimento, penale ESCO se sotto garanzia, ripartizione nello shared savings.
- Report ESG multi-vettoreAggregazione per vettore via API
Energia primaria in kWh, emissioni in kgCO2e, acqua e costi per periodo e insieme di impianti, applicando i fattori di conversione.
- Feed BI ODataOData
Collega Power BI ed Excel direttamente al dato di produzione (impianti, sensori, ordini di lavoro), con autenticazione per tenant.
- Master data export verso ERPJSON (outbox/webhook)
Esporta i dataset impianti/asset e ricambi su configurazioni schedulate con mappatura dei campi, per allineare l'anagrafica dell'ERP.
Try it
Frequently asked questions
Che cosa misura esattamente?
Voltage and current measured by the nodes in the field, from which power and power factor are derived. The values are sent to a central server that stores, aggregates, and displays them on dashboards, with thresholds and alarms.
Da quali sistemi può leggere i dati?
In addition to its own measurement nodes, PlantPulse queries existing sources within the plant via Modbus TCP, OPC-UA, and BACnet/IP. Often, the data needed is already available, and the task is not to install sensors but to read them.
Che cosa succede se la rete cade?
The field agent accumulates readings and sends them back when the connection is restored. A network outage becomes a delay in data, not a gap: it's the difference between a graph with a permanent hole and one that repairs itself.
C'è un'applicazione mobile?
The interface is web-based and adapts to the phone screen. It is a declared choice: two native applications to maintain to display the same graphs would be cost without benefit.
Use cases
Manutenzione predittiva su linea produttiva
I nodi edge misurano la firma elettrica di ciascun asset. La deviazione progressiva viene intercettata dal modulo di salute predittiva, che apre automaticamente un ordine di lavoro con il ricambio gia' associato, prima che il guasto fermi la linea. I piani passano dal calendario fisso al contatore d'uso reale in kWh e ore.
Contratto EPC di una ESCO su portafoglio multi-sito
Una ESCO gestisce piu' clienti sulla stessa installazione. Dove esistono gia' contatori Modbus o PLC OPC-UA il collegamento avviene via gateway, senza installare hardware proprietario. Per ogni sito si definisce la baseline a regressione, i gradi-giorno arrivano dal meteo, e ogni mese il report M&V firmato e il rendiconto EPC raggiungono il cliente sul portale.