How AI Agents and Robotic Arms Are Building Fully Automated Laboratories

Sheena Barazandeh watched in amazement as a robotic arm glided back and forth across the team’s laboratory. The arm loaded a multiwell plate into a device that filled its wells with liquid, then carried the plate to a remote instrument that analyzed the contents. The equipment interacted seamlessly without human intervention.

This autonomous collaboration is powered by a software framework called Model Hardware Standard (MHS). The system connects laboratory instruments from different manufacturers and allows artificial-intelligence systems known as agents to control equipment and coordinate experiments.

MHS is the result of a collaboration between the American artificial-intelligence company Anthropic and the Howard Hughes Medical Institute’s Janelia Research Campus in Ashburn, Virginia.

Why AI laboratory automation is difficult

The platform is designed to address a common problem in scientific research: enabling laboratory devices to communicate with one another. This is notoriously difficult because instruments often come from different vendors and use different programming languages.

MHS can be integrated into any equipment with a programmable interface, allowing it to coordinate communication between devices and support automated experiments.

MHS is not the first tool designed to connect scientific instruments. The SiLA Consortium, a global nonprofit organization that develops frameworks for sharing laboratory data, has created a system called Standardization for Laboratory Automation. It provides a standard language that allows laboratory equipment to communicate with one another.

Unlike MHS, however, the SiLA system does not connect equipment directly to an AI agent.

Setting up experiments in hours instead of months

It “easily takes months” for a research project to move from an initial idea to a practical experiment, says Carnegie Mellon University computational biologist José Lugo-Martinez. Barazandeh, a student on the team, is helping test MHS.

Using the framework, the team set up its experiment in just a few hours. “Being able to reduce that time is great for us,” Lugo-Martinez said.

How the Model Hardware Standard connects lab equipment

To connect devices in a laboratory, scientists and automation engineers typically write custom code that translates between each machine’s programming language. They may also need to add specialized equipment to connect instruments.

“MHS includes software that acts as the ‘connective tissue’ to link the lab computer’s operating system to the instrument, eliminating the need for a custom solution,” said Alec Kemeny, a member of Anthropic’s technical staff and leader of the MHS effort.

Devices connected through MHS can communicate with other connected devices, creating a coordinated system for robotic laboratory automation and AI-controlled experiments.