---
title: 'University of Liverpool: Mobile robotic chemist with Bayesian experiment selection'
slug: liverpool-mobile-robotic-chemist
stable_id: 535dcf1aea9a1411
company: University of Liverpool
function_code: engineering
pattern_codes:
  - autonomous_with_backstop
evidence_strength: verified
publication_tier: showcase
freshness: current
reviewed_at: '2026-08-23'
updated_at: '2026-08-22'
source_quality_summary: 1 peer reviewed, 1 primary; publication outcomes are verified.
caveat_summary: Laboratory proof of concept, not a production deployment; 11 results were discarded.
collections: []
bundle_version: 1.0.0
bundle_fingerprint: sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93
canonical_url: https://brianletort.ai/transformations/liverpool-mobile-robotic-chemist
---

# University of Liverpool: Mobile robotic chemist with Bayesian experiment selection

A robot can automate researcher-facing lab work when search and safety are bounded.

Function: Engineering design. Patterns: Autonomous with a backstop. Evidence: verified.

Freshness: current. Reviewed: 2026-08-23. Updated: 2026-08-22.


Source quality: 1 peer reviewed, 1 primary; publication outcomes are verified.

## Before

1. **Human chemist** — Chooses formulations, prepares samples, operates instruments, records results, and decides the next experiment. (control: Laboratory safety and researcher judgment; authors describe manual search as constrained but provide no measured manual baseline.)
2. **Human researcher** — Interprets results and updates the experimental plan. (control: Manual trial-and-error across a ten-variable space.)

## After

1. **Batched Bayesian search algorithm** — Selects the next batch from prior measured outcomes. (control: Ten-variable bounded search; oxygen and workflow checks discard faulty results.)
2. **Mobile robot** — Navigates the lab, dispenses materials, runs photolysis and chromatography, and returns measurements. (control: Collaborative-robot safety standards; 11 results discarded for workflow errors or high oxygen.)

## Decision rights

Researchers set the problem and safety envelope; the system selects and executes experiments within it.

## Exception path

Faulty seals, high oxygen, workflow errors, or hardware issues are flagged/discarded for human attention.

## Outcomes

- **Autonomous search** (verified): Initial catalyst formulations → 688 experiments in eight days; best mixture six times more active. Laboratory proof of concept, not a production deployment; 11 results were discarded.

## Executive lesson

The operating-model hinge is linking experiment choice, physical execution, measurement, and fault handling.

## Anti-pattern

Do not generalize one autonomous experiment to unsupervised general chemistry.

## Questions for leaders

- Where is the operating threshold set and who can override it?
- What measured result would trigger rollback or retraining?
- Which residual decisions must remain human-owned?

## Sources

- [A mobile robotic chemist](https://www.nature.com/articles/s41586-020-2442-2) — Nature
- [Experiment too big? Hire a mobile robot scientist](https://news.liverpool.ac.uk/2020/07/08/experiment-too-big-hire-a-mobile-robot-scientist/) — University of Liverpool