In recognition of National Science Week, we asked three of our experts to weigh in on the practical uses of robots and what the future of AI might look like.
It’s been just over 100 years since the word “robot” first appeared.
Deriving from the Czech noun “robota”, meaning “forced labour”, the word was first used in a 1921 theatrical play titled R. U. R. (Rossum’s Universal Robots), set in a factory that manufactures artificial people who are designed to serve humans and work for them.
Eventually, the robots turn on their masters, and in the process, wipe out the entire human race.
Sound familiar? It’s a theme that has been exploited many times since in books, TV, movies and in pop culture.
Now, just one century after the word was coined, the idea of robotics is no longer relegated to the realms of futuristic, far-fetched science fiction. Robots have well and truly become entrenched in our world, helping us humanoids in multiple ways, with the technology that drives Artificial Intelligence (AI) advancing every day.
So, what are some of the practical uses of robots? What might the future of AI look like? And what are the opportunities for students to get involved?
We asked three of our experts to give us a byte-size answer.
Dr Christofer J. Clemente, UniSC Associate Professor in Animal Ecophysiology
Q: Many robo-engineers are looking to animals as a source of inspiration, including your team that studied lizard movement to create a climbing robot. What are some of the practical uses for robots that mimic animals?
A: There are two great uses for animal-mimicking robots. Firstly, we can use animals as bioinspiration to build more agile climbing and running robots, to access areas following natural disasters and look for survivors, or use them for maintenance in places that are hard to reach, like the hulls of ships or spacecraft.
But a second exciting use, is we can use robots to better understand biology. By building robots which resemble animals we can see how changes in shape or movement affect the speed or stability of animals, which allows us to understand why some animals have evolved into the diversity or forms we see today.
Dr Umer Izhar, UniSC Lecturer in Mechatronics Engineering
Q: What are the opportunities for our students to get involved in the exciting field of robotics?
A: Our students are learning to plan and code tasks with mobile robots, drones, and robot arms which are relevant to real-world scenarios. They learn to implement machine vision and path planning techniques on different robots which can be used during visual surveying.
We work with sensor fusion techniques to analyse data from multiple sources which can improve time to task for robots in warehousing and packaging automation industries. We are also passionate about contributing to the social engagement experience of robots with humans through assistive robotics.
Dr Erica Mealy, UniSC Lecturer in Computer Science
Q: When it comes to AI in robots, what might the future look like?
A: The future of AI in robots holds the promise of many great things: ensuring that humans don’t need to endanger their lives in risky professions like mining, supporting our aging population to live independently in their own homes for much longer, off-loading mundane and repetitive everyday tasks like carrying loads and way-finding amongst many other possibilities.
For me the issue isn’t what AI can do in robotics but what *SHOULD* it do? How do we manage, design and regulate it? Current AI technologies are divided into classifying and generative. An example of classifying AI is human avoidance for autonomous cars and robots. Classifying AI is already able to accomplish tasks like use lidar sensing to detect, grab and open the door to an assisted living style toilet.
The larger worry is generative AI like chatGPT – it’s accuracy is variable, and they are starting to be used in tasks they are not appropriate for. Just this month Stanford University released a study showing the accuracy of a math task completed by GPT4 reduced from 97.6% in March 2023 to 2.4% in June 2023 - this variability is something we’re not accustomed to with our software and robots. Our tools, whether mechanical, software or robotic, either work or they don’t – or at worst, as in mechanical engineering, degrade over a known period according to the laws of physics and materials science.
The future with AI is already here, but how we manage it is up to us and our regulators! I’m all for removing repetitive, dangerous or mundane tasks and releasing humans to the greater complexities and joys of life, but we must keep an eye on the potential cost if we deploy first and consider later.
Become part of the future of AI and robotics. Start by exploring your study options.
Media enquiries: Please contact the Media Team media@usc.edu.au