ΑΙhub.org
 

From sketches to a robot with artificial intelligence

by
08 July 2024



share this:

PhD-student Elena Jiménez Romanillo with the Master´s students working on a robot with assistance from generative AI. Photo credit: Thor Balkhed.

By Ulrik Svedin

How do you develop a product with as little human involvement as possible? Linköping University students built a robot using generative artificial intelligence.

From an initial idea with AI-generated images to the final stages of optimisation, generative AI (GAI) supported the students throughout the design process.

“I mainly used ChatGPT to program the robot’s navigation and control,” says Arad Jafari, master’s student in technical design.

Programming

The robot is similar to a radio-controlled car, with a small arm and a grapple claw in the front. With the sensors, it is able to drive along a lined path and stop in the designated place, pick up a small box and transport it further.

Master’s student Arad Jafari. Photo credit: Thor Balkhed.

“I’ve always been interested in programming, but my knowledge has been limited. Both my interest and my skills have increased with this project. Robotics and coding were new to me. In addition to the help I have received from the AI, ChatGPT has also acted as a digital mentor,” says Arad Jafari.

How did that work in practice?

“For instance, we received AI support with brainstorming and various solutions to the requirements we have on the robot. What electronic components are needed? How do the components work and how should they be connected? Here, ChatGPT has helped generate all the code for the different solutions.”

The task they faced on the course ‘Design and production’ was to develop a so-called proof of concept, i.e. to prove that something is feasible, in this case by designing, constructing and programming using generative AI – and without too much prior knowledge.

The students´ curiosity

The students were supported by a doctoral student and an external supervisor from the forklift manufacturer Toyota Material Handling. Their supervisor Tomas Jankauskas, senior designer, explains why the industry needs to cooperate with LiU: “We risk turning into dinosaurs if we don’t take advantage of the students’ curiosity. They represent the future. I’ve given them a framework and provided advice, but the students have been the driving force in this project,” he says.

By asking questions to the Open AI tools and a number of other tools (such as ChatGPT, Vizcom and others), they have designed a vehicle with an open chassis in futuristic style. With the help of AI, they have also received assembly instructions. To attach the chassis, they had to make small adjustments manually. But they have basically gone from AI-generated images and 3D models on the screen to 3D printing on the spot. It was difficult in the beginning.

Challenges

“There were products that had three wheels on one side and two wheels on the other. Or an arm that didn’t reach over the hood. A lot that didn’t work. The advantage is that it’s a quick process, so you can fail several times. The difficulty lies in deciding what to move forward with,” says Vivaldus Berglund.

Do you have any tips on how to write questions to an AI?

“We saw that you can ask the AI how best to type a prompt to achieve a certain result. This made the final prompt much better. Another tip is to be specific. The AI will look up shapes based on what it thinks we mean,” says Ville Jonsson.

Why is it important to try to get AI to do the whole job?

“Drawing and programming takes time and costs money. If a manufacturer discovers that it’s possible to automate, that’s of course interesting,” says Johan Asp.

But doctoral student Elena Jiménez Romanillos objects:
“You need to monitor the whole process very carefully and reflect. You have to understand what is happening all the time and spot the mistakes. Otherwise, you’ll lose all the efficiency improvements that can occur with AI,” she says.

PhD-student Elena Jiménez Romanillos. Photo credit: Thor Balkhed

At the same time, the project will be turned into research. Associate Professor Torbjörn Andersson and Elena Jiménez Romanillos study methods for learning using AI. They note that these tools are changing the way students approach different design challenges.

“It’s about how you take facts into account, see connections, draw your own conclusions, evaluate and finally produce and create something new” says Elena Jiménez Romanillos. “For example, the use of AI varies depending on the students’ prior knowledge,”

“Artificial intelligence in the learning process is still a new thing. It’s interesting to see how students search for information and discover completely new tools, almost day by day,” says Torbjörn Andersson.

Exhibition

The students and teachers participated in Milan Design week, 15–21 April 2024, together with Toyota’s exhibitors. They promoted the MSc Programme in Design and Product Development, and shared their experiences of programming and designing using generative artificial intelligence.




Linköping University




            AIhub is supported by:


Related posts :



ChatGPT is changing the way we write. Here’s how – and why it’s a problem

Have you noticed certain words and phrases popping up everywhere lately?
07 October 2024, by

Will humans accept robots that can lie? Scientists find it depends on the lie

Humans don’t just lie to deceive: sometimes we lie to avoid hurting others, breaking one social norm to uphold another.
04 October 2024, by

Explainable AI for detecting and monitoring infrastructure defects

A team of researchers has demonstrated the feasibility of an AI-driven method for crack detection, growth and monitoring.
03 October 2024, by

The Good Robot podcast: the EU AI Act part 2, with Amba Kak and Sarah Myers West from AI NOW

In the second instalment of their EU AI Act series, Eleanor and Kerry talk to Amba Kak and Sarah Myers West
02 October 2024, by

Forthcoming machine learning and AI seminars: October 2024 edition

A list of free-to-attend AI-related seminars that are scheduled to take place between 1 October and 30 November 2024.
01 October 2024, by

Linguistic bias in ChatGPT: Language models reinforce dialect discrimination

Examining how ChatGPT’s behavior changes in response to text in different varieties of English.
30 September 2024, by




AIhub is supported by:






©2024 - Association for the Understanding of Artificial Intelligence


 












©2021 - ROBOTS Association