Integrating Vision and Olfaction via Multi-Modal LLM for Robotic Odor Source Localization
2024

Combining Vision and Smell for Robot Odor Detection

Sample size: 128 publication 10 minutes Evidence: high

Author Information

Author(s): Hassan Sunzid, Wang Lingxiao, Mahmud Khan Raqib

Primary Institution: Louisiana Tech University

Hypothesis

Can integrating vision and olfaction improve robotic odor source localization in complex environments?

Conclusion

The proposed algorithm successfully outperformed traditional methods in locating odor sources in both unidirectional and non-unidirectional airflow environments.

Supporting Evidence

  • The proposed algorithm achieved a 100% success rate in unidirectional airflow environments.
  • In non-unidirectional airflow environments, the proposed method outperformed the Fusion algorithm in terms of success rate and search time.
  • The integration of vision and olfaction allowed for better navigation in complex environments.

Takeaway

This study shows that robots can find smells better when they use both their eyes and noses together, especially in tricky situations.

Methodology

The study implemented a multi-modal LLM-based navigation algorithm on a mobile robot, testing its performance in real-world environments with various airflow conditions.

Limitations

The inference time of the LLM is three seconds, and the experiments were conducted in a small-scale environment.

Statistical Information

P-Value

p<0.05

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.3390/s24247875

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication