1LabTeam

Orchestrating Autonomous Agents for High-Availability Workflows in Industry 6.0

Comment

1LabTeam: Multi-Agent Coordination for Industry 6.0 Workflows

Inspiration

Modern laboratory and industrial workflows remain siloed, manual, and hard to scale. While individual automation tools like lab robots or 3D printers exist, orchestrating them into intelligent, autonomous systems is still a major challenge.

We were inspired by the vision of Industry 6.0—a future where autonomous AI agents coordinate real-world physical and digital processes. Our goal was to explore how Google's Agent Development Kit (ADK) could be used to simulate such a system through a lab-based workflow model.

What it does

1LabTeam is an autonomous, multi-agent orchestration platform built with ADK that simulates a complex laboratory workflow. The system includes:

  • Atlas: A central coordinator agent managing workflow progression
  • Curie: A Bayesian optimizer for experiment parameters
  • Deng: A protocol generator that writes instructions for lab robots
  • Edison: A 3D designer that outputs lab equipment models

Users can trigger the full workflow via a custom UI. A real-time PipelineStatus dashboard and embedded Dev UI provide transparent insights into agent actions, thought processes, and state transitions.

How we built it

  • ADK (Python): Core agent logic and orchestration
  • Vite + React: Frontend interface with workflow visualization
  • Dev UI iframe: Embedded agent trace visualizer
  • Docker-ready architecture: Backend supports unified deployment
  • Optional cloud deployment: Designed for Google Cloud Run

The frontend and backend communicate over local or cloud API, and the system supports plug-in extensions for industrial robotics and 3D printing.

Challenges we ran into

  • Understanding ADK orchestration: Designing proper session, event, and agent transitions took time.
  • Cross-origin iframe integration: We had to build a flexible architecture to serve Dev UI and frontend under a unified host.
  • Multi-agent sequencing: Ensuring proper order and output relay between agents required fine-tuned session management.

Accomplishments that we're proud of

  • Successfully built a fully functional, multi-agent lab simulation with a modular design.
  • Created a custom dashboard (PipelineStatus) with clear execution states.
  • Enabled real-time insight into agent reasoning via Dev UI embedding.
  • Demonstrated how ADK can connect with physical systems like 3D printing and robotic protocol generation.

What we learned

  • How to design collaborative AI agent workflows using ADK.
  • How to bridge frontend UX with backend orchestration in a flexible and extensible way.
  • The architectural patterns needed for real-time, transparent agent-based systems.
  • The potential of ADK as a general-purpose coordination engine across industries.

What's next for 1LabTeam

  • Cloud deployment: Host the full stack via Cloud Run to support public testing.
  • Real hardware integration: Link the Deng agent with physical lab robots (like Opentrons) and Edison with 3D printers.
  • Expand agent library: Add agents for data analytics, sample management, compliance checking, etc.
  • Generalize the platform: Adapt the system for biotech, manufacturing, and supply chain orchestration.

1LabTeam is more than a demo—it's a blueprint for intelligent, autonomous teams working across the physical and digital world.

Built With

Share this project:

Updates