Passing on R&D expertise

Extract researchers' tacit knowledge through AI dialogue and share and leverage it across the organization

Background and Challenges

Research notes and verbal sharing are the norm, and knowledge remains siloed with individuals (person-dependent)

There is a tendency to report only successful cases, resulting in insufficient sharing of failure-related knowledge

Knowledge is dispersed and difficult to search, leading to repeated failures (reinventing the wheel)

AS-IS (Conventional)

・Person-dependent experiment notes

・Insufficient sharing of buried failure knowledge

・Reinventing the wheel (duplicate experiments)

TO-BE (After Implementation)

・Clear records immediately after experiments

・Automatic tagging of failure factors

・Structuring and reuse through a knowledge graph

Operational Flow

1. Input into the system upon completion of experiments and reviews

2. Follow up with interviews for each process to determine whether there were any key points, and record them in the remarks field

3. Integrated into the knowledge graph and linked to related knowledge

4. Recommend past cases when planning similar experiments

Information That Can Be Captured

・Experimental conditions and prerequisite parameters

・Root cause analysis of failures

・Candidate next hypotheses

Implementation Effects (KPI)

・Reduction in duplicate experiments: Significant reduction

・Experiment reproducibility: Improvement

・Exploration lead time: Shortened