Skip to content

What is IvoryOS

IvoryOS turns the Python code that controls your instruments into an app. It reads each instrument’s methods, their arguments and their types, and builds the interface from them. There is no interface code to write and no configuration file to fill in: a scientist designs and runs experiments by pointing and clicking, and anyone who can write a Python class can bring a new instrument.

It sits between point-and-click lab software and writing your own scripts: as easy to use as the first, as open as the second. It runs on the computer next to your instruments and keeps working without an internet connection.

What it isWhere it runs
DeckOne set of instruments with its own workflows and data. The app you spend your day in: Designer, Run, Data History, Instruments, Safety.On the computer beside the hardware (the edge server, a Python process).
Desktop appStarts and watches your decks, installs drivers, and shows every deck’s page in one window.Your computer (Windows, macOS, Linux).
Automation HubA public catalog of instrument drivers, whole platforms, plugins and workflow templates, installable from inside the app.ivoryos.ai
Cloud (early access)Watches and dispatches workflows to the decks in every lab from one place.A web app, hosted or self-hosted.

If you only ever run one bench, the deck and the desktop app are all you need. This guide is mostly about them.

  1. You bring instruments: a Python class per device, written by you, a vendor, or found on the Hub. IvoryOS introspects each class at startup, so every public method becomes a step with a form built from its type hints.
  2. In the Designer you drag those steps into a workflow, fill in values, or leave a value open as a #variable to decide at run time.
  3. You run it Once with the values typed in, Iterate it over a spreadsheet of samples, or Optimize it with Bayesian optimization over the open values.
  4. The deck runs the steps one at a time, pauses for you when a step fails or asks a question, and records every step and value in Data History.
  • The user guide is for the people who run experiments. It walks through each page of the app with screenshots and step-by-step instructions. Start with Try the example.
  • The developer guide is for the people who bring instruments or extend IvoryOS: what it reads from a Python class and what it hides, how to decide what to expose, how to add safeguards, plugins and agents, and how to work on IvoryOS itself. Start with Write a driver.