How to build your own robot friend

However, this technological evolution brings forth significant ethical and equity challenges, prompting questions about universal accessibility to such innovations.

Responding to these concerns, researchers at the University of Southern California (USC) have introduced an innovative, cost-effective solution aimed at democratising AI education. The team has developed a low-cost learning kit that enables college and school students to construct and personalise their own ‘robot friend’. This project not only makes AI more accessible but also incorporates lessons on AI ethics and fairness, ensuring a comprehensive educational experience.

"We’re proposing this open-source model to not only improve education in AI for all students but also to make human-interaction research more affordable for labs and research institutions," stated Zhonghao Shi, a doctoral student at USC's Interaction Lab.

"Ultimately, we want to increase access to human-centred AI education for college students and create a pathway to more accessible research."

To streamline the development process and minimise costs for learners, the USC team adapted Blossom – a small, open-source robot initially developed by Professor Guy Hoffman at Cornell University. Blossom is already a staple in USC’s Interaction Lab, where it has been employed in various projects, including Shi's work on AI voices for mindfulness exercises and colleague Amy O’Connell's adaptation of the robot as a study aid for students with ADHD.

Last year, Shi and O’Connell embarked on redefining Blossom for educational purposes, aiming to create a low-cost, ‘human-focused’ module that mirrors real-world technological interactions.

Their efforts culminated in a new study titled ‘Build Your Own Robot Friend: An Open-Source Learning Module for Accessible and Engaging AI Education’, which was presented at the AAAI Conference on Artificial Intelligence.

"We believe it is important for students to learn about fairness and ethics in AI in the same way that we learned about math and physics in K-12," explained Shi. "We may not use these subjects every day, but having a basic understanding of these concepts helps us do better work and be mindful of new technologies."

The module's affordability is significantly enhanced by using materials producible with 3D printers, substantially reducing costs compared to traditional methods like laser printing. A customisable Blossom robot now costs approximately $250, a stark contrast to the $15,000 price tag of more advanced pre-built robots like NAO.

O’Connell, who took up crochet during the pandemic, has designed five new Blossom exteriors, providing detailed patterns and tutorials for each, including options in knitted and crocheted forms, which further reduce costs and enhance customisability.

The learning module was tested in a two-day workshop in May 2023, involving undergraduate students from a local minority-serving institution. Feedback from the workshop was overwhelmingly positive, with 92% of participants affirming that the experience enhanced their understanding of AI and encouraged them to pursue further studies in robotics and AI.

"Equipping users with AI literacy, including an understanding of AI ethics and fairness, is crucial to avoid unintended discrimination against marginalised groups," Shi added.

As the team continues to refine and expand the module, they aim to extend its reach to younger students and various educational levels, hoping to globally enhance AI education and inclusivity.

"We’re excited to share more about our project with people from around the world," concluded Shi. "We want to make sure that people from different socioeconomic backgrounds have the opportunity to gain an education on AI and participate in the process of improving AI for future use."

This article originally appeared in the May'24 magazine issue of Electronic Specifier Design – see ES's Magazine Archives for more featured publications.