Web HMI and remote machine operation
OEM after-sales remote intervention −68 % · No VPN, no native clientOne responsive web for operator, after-sales and end-customer engineer. Every change signed, no native client, no firewall ports to open.
ISIGECO works as an end-to-end technology partner. We build the artificial intelligence that saves energy, anticipates failures and reorders your stock, and the blockchain chains that give your product provable traceability, from the field or the reactor to the consumer. One team for the firmware, the integration, the agent and the contract.
Well-designed AI agents are not demos: they are silent operators embedded in the 's machine or the end customer's plant, making every minute the micro-decisions a human team would never have time to fine-tune. Remote operation from a single web , machine-as-a-service with settlement, energy savings measured at the electrical cabinet, autonomous material and spares ordering from the machine itself, sensing wired to the HMI, predictive maintenance as an OEM after-sales service, assistants embedded over the OEM manual and regulatory documentation (CE, RoHS, Digital Product Passport) generated on its own. All in production, all in strict compliance with the GDPR and EU Regulation 2024/1689 on Artificial Intelligence.
One responsive web for operator, after-sales and end-customer engineer. Every change signed, no native client, no firewall ports to open.
Your machine charges for its own use. Hours, cycles or kg produced signed at source by the equipment. The gets paid for real use — no manual invoicing, no measurement disputes.
An on-machine agent learns the real consumption profile and retunes setpoints, compressor cycles, heaters and motors without touching the or altering the process recipe.
Every machine knows when its filter, lubricant or electrode is about to run out — and fires the purchase order to the or end-customer ERP. No operator, no email, no paper.
Temperature, humidity, weight, pressure, flow and vibration read by agents that correlate root causes and either fire a action or raise a prioritised alert in the operator's web .
Automatic scoring, tailored nurturing, alert to the human rep at the exact moment, and cross-sell calibrated by channel and customer type.
Time-series signed on the machine + anomaly models trained per-asset. Prioritised alert in the operator's web and in the 's fleet dashboard.
The operator asks in natural language about alarms, parameters or procedures — and gets an answer with an exact citation to the manual and the machine's own log.
Assistants that qualify technical leads and deliver real-time stock, lead time and compatibility.
Production logs, maintenance reports, batch certificates and machine-specific manuals generated from the equipment's own operation. CE, RoHS and EU Machinery Regulation ready without manual work.
GDPR is not an obstacle: it is the strongest contractual guarantee any AI vendor can offer in Europe. We design every system with privacy by design and by default (Art. 25 GDPR), explicitly classifying each use case against the four risk tiers of the EU AI Act — minimal, limited, high and unacceptable — and applying the safeguards each tier demands.
For limited-risk systems (informational chatbots, catalogue assistants) we guarantee algorithmic transparency and the right not to be subject to fully automated decisions (Art. 22 GDPR). For high-risk systems — any AI that drives safety-critical processes or affects fundamental rights — we produce Data Protection Impact Assessments (DPIAs), implement meaningful human oversight and keep audit logs of every decision.
Where technically possible, data is processed inside the European Economic Area. When a component must run outside the EEA, the transfer is covered by the Standard Contractual Clauses approved by Commission Decision (EU) 2021/914, backed by a documented risk assessment. Traceability and auditability of agent decisions are baked in from day one, not bolted on after an audit.
Data minimisation, pseudonymisation by default, justified retention.
Practical explainability and a right to human review wherever it matters.
Processing inside the EEA whenever possible, documented SCCs when not.
Prompts, responses and agent actions recorded for regulatory audit.
Once signed on-chain, plant data and machine data stop being database rows the admin can quietly rewrite: they become mathematically verifiable historical facts — the foundation that new European regulation ( 2024/1781 Digital Product Passport, Machinery Regulation 2023/1230, Asset Administration Shell, W3C Verifiable Credentials 2.0) uses to build industrial digital trust. At ISIGECO we deploy it where value really exists: regulatory compliance (, EMA, AESAN), agrifood traceability, machine digital identity, critical-environment access control, energy certification, water management and Equipment-as-a-Service with pay-per-use. We pick the network around the case — never the other way round.
Public distributed ledger with enterprise governance (Google, IBM, Boeing, Deutsche Telekom). Deterministic 3-5 s finality, negligible per-transaction cost, ~10,000 transactions per second.
The most widespread industrial standard. Solidity, mature tooling (Foundry, Hardhat) and cross-compatibility with any chain (L1 and L2s like Polygon or Base).
Rust + Anchor framework. Up to 65,000 transactions per second, sub-cent fees. The fit when traceability must run at production-line speed.
Extended UTXO model and formal verification with Haskell/Plutus. The choice when mathematical correctness matters more than raw speed.
Every case carries years of regulation behind it. We don't reinvent compliance; we make it auditable at sane cost, with the distributed-ledger network that fits each sector's volume, latency and privacy requirements.
Electronic records and signatures compliant with and EMA Annex 11 , anchored on blockchain for data integrity, untampered audit trail and non-repudiation throughout the batch lifecycle.
End-to-end chain of custody from orchard, fish market or slaughterhouse to the consumer, with per-lot QR and public verification of route, temperature and certifications.
Decentralised identity () + verifiable credentials so only personnel with current training, vaccinations and clearance get in.
Flow, pH, conductivity and pollutant load readings signed at source and immutably recorded. Operator and regulator share the same truth without needing to trust each other.
Guarantees of Origin (GoO/I-REC) issued as tamper-proof digital certificates in real time from the inverter's measured output, not estimated. Automatic settlement of solar surplus between neighbours, intermediary-free.
The business model already running at TRUMPF (Pay-per-Part) and Kaeser (Sigma Air Utility): the designs, installs, operates and maintains; the customer pays per produced unit or m³ delivered. Settlement signed by the machine itself.
Not a buzzword list for a résumé. This is the real stack we ship critical, on-site projects with. Each entry honestly reflects our level and the use cases we have actually run it in.
WCAG 2.1 AA semantic foundation with Schema.org, ready for technical SEO and embedded browsers.
Tokenised design systems with custom properties and modular SCSS. No heavy frameworks.
Modern JS with real-time REST, WebSocket, MQTT and OPC-UA integrations.
TypeScript by default for critical projects. Contracts shared with backend and firmware.
SPA, SSR and micro-frontends with Suspense, Server Components, Zustand and React Query.
Build tool with code splitting, PWA, SEO prerender and per-environment builds.
NestJS backends with guards, interceptors, WebSocket gateways and MQTT.
Industrial ETL, ML with scikit-learn/pandas and hardware-in-the-loop testing.
C for IEC 61131-3 embedded blocks. C# for .NET tooling and industrial supervision.
Microservices and IoT gateways with low latency and a tiny footprint.
Safety-critical firmware and Solana contracts with Anchor.
PLC programming is the heart of the connected factory. At ISIGECO we never separate hardware from software — the equipment we integrate is equipment we program ourselves, and the logic we write runs on architectures we know inside out. That dual view lets us see why a control block behaves differently on an OMRON CPU than on a Siemens one, and pick the right architecture for every project.
NJ/NX CPUs in Sysmac Studio. EtherCAT, EtherNet/IP, motion control and SCARA robots.
TIA Portal V17-V19 for S7-300/400/1200/1500. WinCC, Safety, PROFINET/PROFIBUS.
PC-based automation on Visual Studio. Real-time EtherCAT, motion NC/PTP.
The transversal IEC 61131-3 standard. Turck, Wago, Schneider, Phoenix Contact, Beckhoff.
Early-adopter of web-based CODESYS with MCP Server and Virtual Safe Control SIL3.
Five honest steps. No invented methodologies, no acronyms to sell smoke — just the sequence we have validated by delivering real projects on real plants.
We audit the real environment (existing hardware, protocols, network and physical-space constraints). No honest diagnosis, no solid project.
We document the proposed architecture with readable diagrams, reasoned decisions and a realistic milestone plan. No black boxes.
Short iterations with continuous demos: the customer watches the system grow and can change course while changes are still cheap.
We test in staging and, when risk demands, in HIL simulators of the real process before going anywhere near the plant.
Hand-held rollout, training for the customer's team and post-sales support with clear SLAs. We leave systems the operators can actually maintain.
Let's talk, no strings attached. On the first call we make sure we are a fit; if we are not, we will tell you honestly who can help you better.
Talk to our engineering team