Mobile manipulation robot as autonomous lab assistant

Overview

Carnegie Mellon University offers a mobile manipulation robot designed to serve as an autonomous lab assistant, automating routine manual tasks currently performed by laboratory personnel. The robot can fetch chemicals and equipment to prepare experiments, transport plates from one instrument to another, and be programmed for a wide range of additional laboratory duties. It operates as part of the Programmable Cloud Lab, an integrated platform that combines the robotic hardware with lab management software including a user interface, scheduler, execution module, and data capture capabilities.

The core value proposition is freeing scientists from repetitive, error-prone manual handling so they can focus on research design and analysis. The platform is built from off-the-shelf components and open-source software modules, making it easier to adapt to different laboratories and workflows.

Technical specifications

Key features of the robot include:

• A cylindrical base with a footprint similar to a human, allowing navigation in standard lab environments
• Sensors for navigation, object detection, and experiment observation
• A robotic arm on certain versions for manipulation tasks
• Integration with lab software for user interface, scheduling, execution, and data capture
• Autonomous operation initiated through verbal commands to a language AI

The system was validated in an end-to-end experiment: a user requested a color-mixing experiment verbally, the liquid handler mixed the colors, the robot transported the plate to a spectrometer, and results and logs were saved and presented to the user automatically.

The robot runs on ROS2, an open-source robotics framework, and uses commercially available off-the-shelf hardware, simplifying maintenance, customization, and deployment. The National Robotics Engineering Center (NREC) at Carnegie Mellon University has decades of experience developing robots from concept to commercialization.

Technology readiness level

The robot has completed a basic validation of its core capability: picking up a plate from a liquid handler, transporting it, and placing it in a spectrometer tray within a fully autonomous workflow. A next validation phase is planned to test the robot on real laboratory processes that involve manual handling; test selection, setup, and evaluation criteria will be agreed at the start of the engagement, with results presented through demonstrations and summary materials. The planned duration for this type of validation is a few months. While current validation covers a specific transport workflow, the robots can be programmed for additional tasks, with new skills under development.