Researchers led by Zhengqi Chen found that human fingertips can detect objects buried in sand by sensing tiny grain displacements—a phenomenon called "remote touch." In laboratory trials, humans achieved roughly 71% precision locating hidden polyester cubes, outperforming an LSTM-trained robotic sensor that reached about 40% precision and generated many false positives. The effect depends on granular physics (grain jamming can disrupt cues) and sensitive mechanoreceptors in skin. Authors suggest robotics could benefit from multi-modal sensing and adaptive learning to better mimic human performance.
Hidden Sense in Your Fingertips: Humans Can 'Feel' Buried Objects by Skimming Sand
Researchers led by Zhengqi Chen found that human fingertips can detect objects buried in sand by sensing tiny grain displacements—a phenomenon called "remote touch." In laboratory trials, humans achieved roughly 71% precision locating hidden polyester cubes, outperforming an LSTM-trained robotic sensor that reached about 40% precision and generated many false positives. The effect depends on granular physics (grain jamming can disrupt cues) and sensitive mechanoreceptors in skin. Authors suggest robotics could benefit from multi-modal sensing and adaptive learning to better mimic human performance.
01:49 AM, 11/20/2025Science
Researchers have demonstrated that human fingertips can detect objects buried in granular materials such as sand and gravel by sensing tiny displacements of individual grains. This ability—often called "remote touch"—allows people to infer the presence and approximate location of a hidden item simply by sweeping a finger across the surface, without directly contacting the object itself.
Experiment
In controlled trials led by Zhengqi Chen at Queen Mary University of London, volunteers attempted to locate polyester cubes hidden under dry sand in plastic boxes with opaque lids. For each trial the cube’s position was randomized. Participants inserted an index finger through a hole in the lid, kept pressure constant and matched their scanning speed to a blinking LED strip; they stopped only to indicate where they believed a cube lay.
Findings
Humans performed remarkably well: fingertip remote touch yielded a precision of about 71%. The researchers compared these results with a custom robotic tactile sensor trained with a Long Short-Term Memory (LSTM) algorithm. While the robot sometimes detected objects from a slightly greater range, its precision peaked around 40% and it produced many more false positives than human participants.
How Remote Touch Works
The physics of granular materials helps explain why remote touch succeeds or fails. Sand grains are in continuous contact and transmit stress through inter-particle forces; under some conditions the grains can jam and behave like a solid, disrupting the transmission of subtle forces to the probing finger. Biologically, mechanoreceptors in the skin transduce pressure, vibration and shear forces into neural signals. These receptors are especially sensitive to the small, friction-related motions produced when a finger brushes over a textured or granular surface.
Implications and Future Directions
The study highlights both a little-known human capability and a potential route for improving robotic tactile systems. The authors note that this tactile ability remains largely underexplored despite its relevance to robotic excavation in archaeology, marine and space exploration, and object detection during search-and-rescue operations. They suggest that future robotic systems could reduce false positives by integrating multi-modal sensing and adaptive learning strategies that draw on human perceptual adaptability.
Whether you call it a "sixth sense" or a specialized aspect of touch, Chen and colleagues show that our hands can, in effect, "touch" objects without directly contacting them—an insight that may inspire more sensitive, human-like robots and new minimally invasive search techniques.