Introducing LEAF: A Modular Framework for Equipment Data Collection
Scientific and industrial equipment generates enormous amounts of valuable data and most of it never leaves the device. Whether it is a chromatography system, a sensor array, or a bench-top instrument, the majority of devices lacks the ability to automatically push data to a central environment. That gap is exactly what LEAF has been designed for.
What is LEAF?
LEAF, the Lightweight Equipment Adapter Framework, is an open-source, modular framework that enables automated, standardised, and interoperable data collection from diverse equipment. It gives existing devices Internet of Things capabilities without requiring hardware modifications or changes to manufacturer software.
At its core, LEAF works through adapters: containerised units made up of modules each with its own interpreter. Modules handle the specifics of talking to a piece of equipment (reading registers, polling endpoints, subscribing to streams), while the interpreter transforms that raw data into a standardised structure ready for storage or downstream processing.
Why is this important?
The problem LEAF solves is theoretically simple but hard to solve in practice:
- Equipment speaks dozens of different protocols and data formats
- Each vendor has their own proprietary structure
- Integrating even a handful of devices into a unified pipeline requires significant custom engineering effort
LEAF breaks down those silos. By defining a standard adapter interface, any piece of equipment, regardless of manufacturer or protocol, can feed into the same data infrastructure. Your OPC device, Bluetooth-enabled sensor, or file based device can all output data in the same format, managed by the same system.
How Adapters Work
Adapters are containers of modules that transform equipment data into a standard structure for output or storage. Each module handles communication with a specific piece of equipment — reading registers, polling endpoints, or subscribing to data streams — while the interpreter normalises the raw result into LEAF’s standard format. This separation means the same adapter can feed data to different backends without any changes to the equipment-facing logic.
The Backend: MAPLE (coming soon)
MAPLE is the backend stack that receives, stores, and visualises data from LEAF adapters. It is built around a set of well-established open components:
MAPLE is currently under active development. Documentation and deployment guides will follow once the stack reaches a stable release.
Open and Extensible
LEAF is fully open-source and built to be extended. The adapter marketplace lets teams share and reuse adapters built for specific equipment, avoiding duplicated effort across labs and organisations. If an adapter doesn’t exist for your equipment yet, the framework makes it straightforward to build one.
What’s Next
We’re actively growing the adapter ecosystem and improving the tooling around LEAF. In recent months we’ve added support for several widely-used protocols and have build adapters for various devices. In the coming months, we plan to expand the marketplace, add a web-based management interface, and continue improving the core framework based on user feedback.
If you’re dealing with equipment data silos and want to explore how LEAF can help, check out the documentation or browse the adapter marketplace.