Revealing Mysteries: How Asteroid Grains Illuminate the Birth of the Outer Solar System

Tiny grains from asteroid Ryugu are revealing clues to the magnetic forces that shaped the far reaches of the solar system over 4.6 billion years ago. The findings suggest the distal solar system harbored a weak magnetic field, which could have played a role in forming the giant planets and other objects. Tiny grains from
HomeTechnologyMoo-ving Towards Better Human-Robot Connections: The Role of Virtual Cows

Moo-ving Towards Better Human-Robot Connections: The Role of Virtual Cows

A video game that has players herding virtual cattle has enhanced our understanding of human decision-making in movement and navigation. This insight could improve how we engage with artificial intelligence and refine the movement of robots in the future.

A video game in which participants herded virtual cattle has furthered our understanding of how humans make decisions on movement and navigation, and it could help us not only interact more effectively with artificial intelligence, but even improve the way robots move in the future.

Researchers from Macquarie University in Australia, Scuola Superiore Meridionale, the University of Naples Federico II, the University of Bologna in Italy, and University College London in the UK utilized the video game in their study to uncover how dynamic perceptual-motor primitives (DPMPs) can replicate human decision-making.

DPMPs are mathematical constructs that aid in understanding how we coordinate our movements in response to various stimuli in our environment. They have proven useful in illuminating how we navigate and perform different tasks.

This is especially significant in complex settings featuring other individuals and both stationary and moving objects, such as busy sidewalks or sports fields.

Previously, it was believed that our brains created extensive maps of our surroundings and then devised detailed plans for movement. However, recent studies suggest that instead of following a rigid plan, we move intuitively, considering our objectives and navigating around obstacles we encounter.

The new research, published in the latest issue of Royal Society Open Science, had participants engage in two herding activities where they moved either a single cow or a group into a pen.

The researchers monitored the sequence in which players guided the cows and used that data in their DPMP model to see if it could replicate the actions of the human players.

Lead author and PhD candidate Ayman bin Kamruddin noted that the DPMP model successfully reflected how players moved and predicted their decisions.

Professor Richardson mentioned, “In the multi-target task, three patterns appeared in how participants chose their targets: they tended to select the nearest cow first based on angular distance, each subsequent cow was closest to the last chosen, and if faced with a choice between two cows, they usually picked the one furthest from the center of the pen.”

“After incorporating these three decision-making rules into the DPMP, it could accurately predict nearly 80% of the choices regarding which cows to herd next and also anticipated how participants would behave in new scenarios with multiple cows.”

Herding games are commonly used in research like this due to their resemblance to real-life situations where individuals must control other agents.

Previously conducted studies utilized a top-down view of the animals, raising concerns that this perspective may skew participants’ decisions, as they might behave differently in real-life situations where they lack such a comprehensive overview.

To address this issue, the research team created a new herding game that restricted players’ field of vision to what is typically visible from a first-person perspective, similar to many role-playing video games.

Senior author Professor Michael Richardson from the Macquarie University Performance and Expertise Research Centre states that this shift in perspective has crucial implications.

“While earlier research has demonstrated that DPMPs can predict crowd dynamics or track a moving object, our study is unique in examining whether the model can be applied to how a person directs a virtual character or robot,” he explained.

“This represents a significant move towards developing more responsive and intelligent systems.

“Our findings underscore the need to incorporate intelligent decision-making strategies into DPMP models to enhance how robots and AI emulate human movement, behavior, and interaction.”

“They also suggest that DPMPs could be valuable in real-world applications, such as crowd management, evacuation planning, virtual reality training for firefighters, and even search and rescue operations, as they can help predict human reactions and movement patterns.”