Automata and Anthropic Enable Self-Healing Labs with AI-Powered Recovery

An automation protocol rarely fails when repeated under ideal conditions. Deviations such as loading mistakes or changes in run order are likely culprits. For attended platforms, the errors caused by these deviations can drive a rush back to the workcell to correct a time-sensitive issue. For overnight cell culture runs, they could ruin an experiment and set the lab months behind schedule.

‍

Enter Automata LINQ and MHS

LINQ, Automata’s AI-ready lab automation platform, is designed to use the latest tools available to simplify common problems like error recovery. Using MHS (Model Hardware Standard), Automata has demonstrated automated error recovery on a workcell without user intervention. When labware is left in the wrong slot, LINQ exposes the deviation and calls Claude using MHS to interpret, diagnose, and prescribe the recovery output. LINQ then assesses the prescribed recovery output and replans the run to execute the fix - all planned and deployed without any action required from the operator.

‍

Finding a Missing Plate, Never Missing a Beat

In this example use case, the robotic arm on the LINQ bench goes to retrieve a tube rack from an expected location. The missing rack triggers an error (the point where most schedulers would stop and await user action). LINQ calls Claude to diagnose the issue where the missing plate is confirmed via on-bench cameras. The missing tube rack is found and the error recovery actions are sent to LINQ (pick tube rack from centrifuge, place on scanner, retry scan task). The camera is checked again to ensure the issue is resolved before LINQ resolves the error and continues the run. All of this can be seen in the LINQ Control Room, but requires no intervention. An operator in a meeting would come back to the system running as intended.

‍

Minor Corrections, Major Effects

Simple plate presence detections like the one shown in our example are anything but trivial. Our data shows that roughly 80% of errors associated with the robotic arm are tied to improperly located labware. Each one of these errors would have been unrecoverable without intervention in the past!

Even when instrument or environmental limitations require that a user still resolve the error, such as an incorrectly configured liquid handling deck or a room too hot for an incubator to maintain temperature, LINQ and the MHS connector can identify the root cause and enable a fast resolution by the operator. 

‍

Practical By Design. Intelligent By Intent.

The core focus of intelligent lab automation, powered by LINQ, is to make life in the lab as simple and powerful as possible. With LINQ and MHS working together, users get fewer stalls and higher utilization, leading to self-healing labs that notice anomalies and correct courses on their own. 

That’s what Automata is building towards with LINQ and Anthropic.

‍

Automata is transforming lab automation for the AI era, delivering a fully integrated, AI-ready platform that is modular, configurable, and open by design. While legacy systems fragment workflows and slow discovery, Automata modernizes lab operations, enabling scientists to program experiments, not robots.

For more information, visit www.automata.tech.