A research team from Queen Mary University of London and University College London recently released research results, proposing for the first time that humans may have a previously unrecognized sensory ability - "remote touch", that is, they can sense the existence of a target object through weak mechanical perturbations in the medium before directly contacting it. This discovery is considered to have changed the traditional cognitive framework of human touch as "only close and necessary contact", and is expected to provide new ideas for the design of robotics and various assistive devices.
While human touch has long been viewed as a short-range sense that relies heavily on direct skin-to-object contact, research on how animals perceive the world has gradually produced different evidence. For example, some shorebirds, such as sandpipers and plovers, have been found to be able to locate prey buried beneath the sand by sensing extremely subtle mechanical disturbances in the sand, a process known as long-range touch. Under this mechanism, when objects in the surrounding environment move, small pressure changes or movement reflexes are produced in the granular medium, and animals capture these signals through a highly sensitive tactile system.
The latest research raises the question on this basis: Do humans have similar abilities? The research team reported the relevant experimental design and results at the 2025 IEEE "International Conference on Development and Learning" (ICDL). In the experiment, the subjects were asked to gently slide their fingers across a box filled with sand, and before touching the target object, try to determine whether there was a block buried in a fixed position in the box. The research team controlled the position of the blocks and used LED light strips to guide the subjects' stroking trajectories, thereby systematically recording their judgment performance.
Experimental results showed that human participants were able to sense the presence of buried blocks with significantly better than chance accuracy without directly touching them. This means that when a human finger sweeps across the sand surface, it can actually feel very small changes in the direction or resistance of the sand grains due to the presence of the solid underneath. Further physical modeling analysis pointed out that the sensitivity of this perceptual ability is close to the mechanical reflection detection limit predicted by theory, showing that the human hand tactile system is far more refined than the "rough" in traditional understanding.
In order to compare the performance of humans and machines on remote touch, the team also designed parallel robot experiments. In the robot experiment, the researchers used a UR5 robotic arm equipped with a tactile sensor to also detect buried blocks in the sand, and used a long short-term memory (LSTM) algorithm to learn and judge the sensing data. The results show that within the theoretical detectable range, the human subject's judgment accuracy is about 70.7%. Although the robot has a slight advantage in the average detection distance, the overall accuracy is only 40% due to more false alarms.
The researchers pointed out that the performance of both is very close to the maximum sensitivity limit predicted by the physical model, which confirms the "physical boundary" of remote touch in granular media. What’s more noteworthy is that there is a positive interaction between human and machine experiments: human experiments provide inspiration for machine learning strategies, and machine performance helps researchers interpret human behavior data from a new perspective. The research team believes that this combination of psychology and robotics demonstrates the potential of interdisciplinary collaboration in basic cognitive research and technological innovation.
Chen Zhengqi, the first author of the paper and a doctoral student in the Advanced Robotics Laboratory of Queen Mary University of London, said that this discovery opens up new directions for designing tools and assistive technologies that can expand human tactile capabilities. For example, in the future, sophisticated detection equipment for archaeological excavations can be developed to locate targets buried in sand without destroying cultural relics. The same principle is also expected to be used for gravel exploration on the surface of stars such as Mars, or search and rescue in seafloor sand environments, making it safer and more efficient to carry out "blind exploration" missions in environments with limited vision or even dangerous conditions.
Study leader Elisabetta Versace, senior lecturer in psychology at the Prepared Minds Laboratory at Queen Mary University of London, pointed out that this is the first time that remote touch has been systematically studied on the human body, and it is proposed that the boundaries of humans' "range of perceptual world" (the so-called receptive field) may be far beyond what was previously recognized. She believes that this work not only expands the theoretical understanding of human perceptual systems, but also provides new ideas for how people interact with complex environments.
Lorenzo Jamone, co-author of the paper and associate professor of robotics and artificial intelligence at University College London, emphasized that what is special about this study is that human experiments and robot experiments "borrow" each other. In his view, this kind of cross-field cooperation not only helps researchers identify human beings' yet-to-be-discovered sensory capabilities, but also lays the foundation for the development of robots with human-like tactile "intuition." In the future, with the further development of sensor technology and algorithms, remote touch is expected to become one of the important capabilities for robots to complete tasks in extreme scenes such as darkness, turbidity, and deep burial.
The research is titled "Exploring Tactile Perception for Object Localization in Granular Media: A Human and Robotic Study" and was published in the Proceedings of the 2025 IEEE International Conference on Development and Learning. The study included two core experiments: one focused on assessing the sensitivity of human fingertips to tactile cues of objects buried in the sand, and the other used a haptic-equipped robot arm combined with a long-short-term memory model to determine whether the target exists or not. The research team stated that follow-up work will further explore the performance of remote touch in other media (such as soil or mixtures of different particle sizes) and evaluate the potential of this ability for practical applications in daily life and professional tasks.