Plant data, configured, not coded
For years I have watched plants pay for big SCADA licences and then wait weeks for every small change.
A new machine means new screens. A new alarm means a new script. A new report means another visit from the integrator. And when the auditor arrives, the quality team scrambles to put together records that should have been there all along.
None of this is because the people involved are slow. It is because most plant software is built as a programming project. Every change goes back to someone who can code, and every plant ends up with a system that only one or two people truly understand.
We at Bufferstack.IO wanted to build something different.
The Bufferstack IIoT Gateway connects to your machines over OPC UA, and everything is set up in one configuration file. Machines, tags, alarms, recipes, OEE and reports are declared there. There is no scripting and no screen drawing. If you can read a configuration file, you can add a machine.
It works with Kepware, Mitsubishi PLCs, MTConnect machines and any standard OPC UA server, so it sits on top of what a plant already has.
There is one package, and it includes everything:
- Live dashboards in any browser, with state, telemetry, IO, tables and charts
- OEE per machine, calculated live
- Digital alarms, and four-level analog alarms (HIHI, HI, LO, LOLO) with deadband
- A PostgreSQL historian with time-based and trigger-based logging
- Printed data, alarm and audit trail reports, and CSV export
- SPC control charts with Nelson rules
- Setpoints, buttons and recipes, with every write e-signed and audited
- 21 CFR Part 11 e-signatures, audit trail, role-based access, password ageing and account lockout
We deliberately did not split these into tiers. A plant should not discover at audit time that compliance was an add-on it did not buy.
Knowing a system means knowing how it fails, not only how it works. So we started with the failures.
PLCs lose power. Network switches get rebooted. The database server goes down for maintenance. In a plant these are not rare events; they are a normal week.
The gateway handles each of them on its own. It reconnects to PLCs and rebuilds its subscriptions without anyone touching it. If the database goes down, audit records are written to a local spool and replayed once it is back, so the audit trail has no silent gaps. Each group of machines runs in its own worker process, so one misbehaving server does not take down the rest.
It is also light. The gateway and its PostgreSQL historian run together on a dual-core machine. In our lab load tests, one gateway handled 20,000 tags from a single OPC UA server using a small share of one CPU core.
The gateway is built for pharma and regulated manufacturing, assembly lines that want accurate OEE, machine shops on MTConnect, and plants running Kepware or Mitsubishi PLCs with support for OPCUA connectivity.
I want to be honest about where it stands today. It has been load-tested in our lab with tens of thousands of tags, PLC power cuts, network drops and database restarts. It has not yet run in a production plant. That is the next step, and I want to take it with the right people.
So I am looking for two kinds of early partners:
- System integrators who will train their engineers on it, test it on their own bench with their own PLCs, and then take it to a first customer site.
- Plants willing to be pilot sites, with SI on site during commissioning and direct support afterwards
Early partners will get priority fixes, a real say in what comes next, and pricing that reflects the risk they are taking with a new product.
Let us talk
If you are a system integrator or a plant that wants to see this working on your own machines, I would be glad to hear from you.
Latest from Bufferstack.IO
- Plant data, configured, not coded October 5, 2026For years I have watched plants pay for big SCADA licences and then wait weeks for every small change. A new machine means new screens. A new alarm means a new script. A new report means another visit from the integrator. And when the auditor arrives, the quality team scrambles to put together records that […]harshadjoshi
- Stop Recording Mistakes in Manufacturing. Start Preventing Them. June 5, 2026In manufacturing, “traceability” is usually treated as an afterthought. Most lines rely on fragile SCADA scripts that blindly dump data into flat files. They don’t control the process; they just record what happened – even if what happened was wrong. When operators are rushed, steps get skipped. When parts are scrapped, sub-components mysteriously find their […]harshadjoshi
- We Open-Sourced Our Atlas Copco Open Protocol Library February 17, 2026Ever wondered why connecting to industrial nutrunners (those torque tools that tighten bolts in factories) is harder than it should be? There’s a standard protocol. There’s documentation. There are even a few libraries on GitHub. But try using them in a real factory and things fall apart quickly. Network cable gets unplugged? Connection’s dead. Multi-spindle […]harshadjoshi
- We have something for Mitsubishi FA product users November 10, 2025We’ve been working on a new tool and wanted to share it with you. What We Built?It’s called the Bufferstack.IO MCProtocol OPC UA Server. Basically, it helps connect Mitsubishi PLCs to any software that speaks OPC UA. Why We Made It?A few months back, we wrote about how hard it is to find good libraries […]harshadjoshi
- Why AI/LLM Integration Remains Unfeasible in India’s Industrial Automation Sector – 2025 Reality Check August 25, 2025As we step deeper into 2025, the buzz around integrating advanced technologies like Artificial Intelligence (AI), Large Language Models (LLMs), and Model Context Protocol (MCP) servers into India’s industrial automation sector is louder than ever. MCP servers, for those unfamiliar, act as a bridge enabling LLMs and AI agents to securely interact with real-world tools, […]harshadjoshi
- The Scarcity of Open-Source Libraries for Mitsubishi PLC MC Protocol, SLMP, and CC-Link June 11, 2025Ever wondered why open-source is booming in Information Technology (IT) but remains a rare sight in the industrial Operational Technology (OT) landscape? Our latest analysis, drawing from the “The Scarcity of Open-Source Libraries” report, unpacks this significant divide, especially concerning proprietary Programmable Logic Controller (PLC) protocols from industry giants like Mitsubishi Electric. While a staggering […]harshadjoshi
- Bufferstack.IO OPC UA MTConnect Adapter Development Notes April 4, 2025As software engineers, we often dive into solving problems headfirst, eager to write code and see results. But a recent experience reminded me of the importance of following a proper engineering approach—especially when dealing with complex standards like MTConnect. Here’s a blog post about how we resolved a tricky issue, the time we wasted by […]harshadjoshi
- Introducing Bufferstack.IO Universal OPCUA Server Shareware November 29, 2024We are excited to announce the release of our latest shareware application: the Bufferstack.IO Universal OPCUA Server. This versatile tool is designed for developers, engineers, and learners to explore OPC UA server functionality, integrate with Modbus devices, and evaluate secure industrial communication standards. This shareware version is intended for evaluation purposes only. What is the […]harshadjoshi
- Enhancing Security on Bufferstack.IO Super Container OS with TOTP July 22, 2024How to add totp security feature to Bufferstack.IO Super Container OS and this information can also be used for other #Debian derived distros.harshadjoshi
- Super Container OS Release Notes 14-06-2024 June 14, 2024The Bufferstack.IO team is thrilled to announce the latest iteration of Super Container OS, bringing a wave of advancements for streamlined container orchestration and infrastructure deployment. This release empowers developers and IT professionals with a robust platform to manage their containerized applications with greater ease and efficiency. Enhanced Core: Streamlined Infrastructure as Code: Vagrant is […]harshadjoshi