A Robot Chemist that can Learn from Humans

Chemistry experimentation is fundamental to scientific research and development. Nevertheless, it often entails time-consuming, labour-intensive processes that are susceptible to human error. Robotic systems hold immense promise in transforming this field by automating and streamlining chemistry laboratory experiments.

However, due to the intricate dexterity required, programming the behaviour through explicit instructions can be a formidable challenge. This research endeavour seeks to tap into human expertise and explore the potential of utilising Learning from Demonstrations (LfD) techniques. This approach empowers robots to learn directly from human demonstrations, enabling them to perform chemistry experiments with increased efficiency and precision.

The study will specifically investigate the application of Schlenk line techniques, transferring knowledge and extensive hands-on experience from humans to robot through kinesthetic teaching or teleoperation. Aims and objectives: - Identify key manipulation skills in chemistry "Schlenk line" experiments

- Implement learning from demonstration techniques to enable the robot to acquire and refine these skills through observation and imitation.

- Evaluate the performance and effectiveness of the robotic chemist in comparison to traditional human-led experimentation.

To achieve the stated objectives, a multi-step approach will be undertaken. Firstly, it is imperative to conduct a detailed analysis of the key manipulation skills involved in chemistry "Schlenk line" experiments. Once these skills are identified, the next step would involve implementing learning from demonstration techniques on the Franka Research 3 Robots.

Leveraging their capabilities in torque control and integration of both vision and tactile feedback, the robots can be trained to observe and imitate human operators performing these tasks. This would require collecting a comprehensive training dataset, encompassing various scenarios and conditions that might be encountered in a typical Schlenk line experiment.

Subsequently, a supervised learning approach can be employed to facilitate skill acquisition. The skill will be further refined through reinforcement learning or repeated demonstrations and feedback. Finally, a thorough evaluation of the robotic chemist's performance would be conducted, comparing it to traditional human-led experimentation.

Metrics such as precision, accuracy, efficiency, and safety will be assessed to determine the effectiveness and potential advantages of the robotic system. This comprehensive approach aims to bridge the gap between human expertise and robotic automation in the domain of chemistry experimentation, specifically Schlenk lines which are widely used in synthetic chemistry laboratories around the world.

Alignment to EPSRC's strategies and research areas = artificial intelligence and robotics