Researchers at TUDelft and the Swiss technical university with the help of chatGPT, have developed a tomato-harvesting robot after deliberating and accepting the challenges.
ChatGPT excels in handling poems, essays, books, and potentially designing robots. However, there are risks and benefits associated with using ChatGPT in the design process.
Researchers at TUDelft and the Swiss technical university EPFL with the help of chatGPT, have developed a tomato-harvesting robot. After discussing with ChatGPT, they decided to tackle the challenges and designed a robot.
Helpful suggestions
Researchers adopted ChatGPT’s design decisions, especially valuing its input during the conceptual phase. The researchers highlight that ChatGPT expands designers’ knowledge across various domains, such as identifying the economically valuable crop for automation. During implementation, ChatGPT provided helpful suggestions like “Use silicone or rubber for the gripper to prevent tomato crushing” and “Employ a Dynamixel motor for optimal robot driving.” The collaborative outcome is a tomato-harvesting robotic arm, showcasing the synergy of humans and Artificial intelligence (AI).
ChatGPT as a researcher
The researchers deemed the collaborative design process positive and enriching. However, the role of engineers shifted towards technical tasks. In an exceptionally extreme scenario, the AI takes complete control over the robot design process, providing all the necessary input, while the human follows its lead without question. In this setup, the Large Language Model (LLM) is the researcher and engineer responsible for all technical aspects. At the same time, the human assumes the role of a manager, tasked with defining the design objectives and overseeing the process.
Risk of misinformation
Today’s LLMs cannot achieve such extreme scenarios, and their desirability is questionable. The unverified LLM output can be misleading, carrying risks of misinformation and bias in robotics. Collaborating with LLMs raises concerns about plagiarism, traceability, and intellectual property. The team plan to utilize the tomato-harvesting robot in their ongoing robotics research. They also explore LLMs’ role in designing new robots, mainly focusing on AI autonomy in shaping their bodies. The researchers concluded by emphasizing the open question of using LLMs to assist robot developers while preserving the creativity and innovation necessary to tackle 21st-century challenges in robotics.
Reference: Francesco Stella et al, How can LLMs transform the robotic design process?, Nature Machine Intelligence (2023). DOI: 10.1038/s42256-023-00669-7. www.nature.com/articles/s42256-023-00669-7
