Summary: A new study shows that humans possess a form of “remote touch,” allowing them to detect hidden objects in sand before making direct contact. Participants sensed buried cubes by perceiving tiny mechanical reflections generated as they moved their fingertips through the sand.
Humans performed comparably to physical models of maximum detectability and outperformed an LSTM-trained robotic sensor in precision. The discovery expands our understanding of human tactile perception and provides a blueprint for designing advanced robotic systems capable of delicate, vision-limited tasks.
Key Facts
Human Remote Touch: People can detect buried objects before contact by sensing tiny mechanical disturbances in sand.High Precision: Humans reached ~70% precision, outperforming a robot that detected farther but produced more false positives.Tech Inspiration: Findings offer benchmarks for improving tactile-based robotics, assistive tools, and exploration technologies.
Source: Queen Mary University London
A study by researchers at Queen Mary University of London and University College London has found that humans have a form of remote touch, or the ability to sense objects without direct contact, a sense that some animals have.
Human touch is typically understood as a proximal sense, limited to what we physically touch. However, recent findings in animal sensory systems have challenged this view. Certain shorebirds, such as sandpipers and plovers, use a form of “remote touch” to detect prey hidden beneath the sand (du Toit et al. 2020; de Fouw et al. 2016).
The findings also offer valuable benchmarks for improving assistive technology and robotic tactile sensing. Credit: Neuroscience News
Remote touch allows the detection of objects buried under granular materials through subtle mechanical cues transmitted through the medium, when a moving pressure is applied nearby.
The study in IEEE International Conference on Development and Learning (ICDL) investigated whether humans share a similar capability. Participants moved their fingers gently through sand to locate a hidden cube before physically touching it.
Remarkably, the results revealed a comparable ability to that seen in shorebirds, despite humans lacking the specialized beak structures that enable this sense in birds.
Results show human hands have more sensitivity than expected
By modelling the physical aspects of the phenomenon, the study found that human hands are remarkably sensitive, detecting the presence of buried objects by perceiving minute displacements in the sand surrounding them.
This sensitivity approaches the theoretical physical threshold of what can be detected from mechanical “reflections” in granular material, when there is a sand movement that is “reflected” on a stable surface (the hidden object).
Do humans or robots perform better on remote touch?
When comparing a human’s performance with a robotic tactile sensor trained using a Long Short-Term Memory (LSTM) algorithm, humans achieved an impressive 70.7% precision within the expected detectable range. Interestingly, the robot could sense objects from slightly farther distances on average but often produced false positives, yielding only 40% overall precision.
These findings confirm that people can genuinely sense an object before physical contact, a surprising capacity for a sense that is usually concerned with objects that enter in direct contact with us. Both humans and robots performed very close to the maximum sensitivity predicted with physical models and displacement.
Why is the study important?
This research reveals that humans can detect objects buried in sand before actual contact, expanding our understanding of how far the sense of touch can reach. It provides quantitative evidence for a tactile skill not previously documented in humans.
The findings also offer valuable benchmarks for improving assistive technology and robotic tactile sensing. By using human perception as a model, engineers can design robotic systems that integrate natural-like touch sensitivity for real-world applications such as probing, excavation, or search tasks where vision is limited.
What are the wider implications?
Elisabetta Versace, Senior Lecturer in Psychology and lead of the Prepared Minds Lab at Queen Mary University of London who conceived the human experiments said: “It’s the first time that remote touch has been studied in humans and it changes our conception of the perceptual world (what is called the “receptive field”) in living beings, including humans.”
Zhengqi Chen, PhD student of Advanced Robotics Lab at Queen Mary University of London said: : “The discovery opens possibilities for designing tools and assistive technologies that extend human tactile perception.
“These insights could inform the development of advanced robots capable of delicate operations, for example locating archaeological artifacts without damage, or exploring sandy or granular terrains such as Martian soil or ocean floors.
“More broadly, this research paves the way for touch-based systems that make hidden or hazardous exploration safer, smarter, and more effective.”
Lorenzo Jamone, Associate Professor in Robotics & AI at University College London, said:
“What makes this research especially exciting is how the human and robotic studies informed each other. The human experiments guided the robot’s learning approach, and the robot’s performance provided new perspectives for interpreting the human data.
“It’s a great example of how psychology, robotics, and artificial intelligence can come together, showing that multidisciplinary collaboration can spark both fundamental discoveries and technological innovation.”
The studies:
Researchers carried out two studies: the first, a human study assessing fingertip sensitivity to tactile cues from buried objects; the second, a robotic experiment using a tactile-equipped robotic arm and a Long Short-Term Memory model to detect object presence.
The authors are Zhengqi Chen, PhD student of Advanced Robotics Lab, Dr Laura Crucianelli Lecturer in Psychology, Dr Elisabetta Versace, Senior Lecturer in Psychology, all from Queen Mary University of London and Lorenzo Jamone, Associate Professor in Robotics and AI at University College London.
Key Questions Answered:Q: What did researchers discover about human touch?
A: Humans can detect buried objects in sand before making physical contact, showing a form of “remote touch” previously known only in certain animals.
Q: How does this sensing ability work?
A: Minute mechanical displacements in sand “reflect” off hidden objects, and the human hand can pick up these subtle cues while moving through the material.
Q: How accurate were humans compared to a robotic tactile system?
A: Humans achieved about 70% precision within the detectable range, outperforming a robot that sensed objects slightly farther away but produced many false positives.
About this tactile perception and neuroscience research news
Author: Lucia Graves
Source: Queen Mary University London
Contact: Lucia Graves – Queen Mary University London
Image: The image is credited to Neuroscience News
Original Research: Closed access.
“Exploring Tactile Perception for Object Localization in Granular Media: A Human and Robotic Study” by Elisabetta Versace et al. IEEE International Conference on Development and Learning
Abstract
Exploring Tactile Perception for Object Localization in Granular Media: A Human and Robotic Study
Localizing buried objects in granular media, such as sand, poses a significant challenge in robotics, with limited insight into human tactile capabilities in these environments.
This study presents a novel approach to tactile-based localization through: (1) a human study with 12 participants assessing fingertip sensitivity to tactile cues from buried objects, and (2) a robotic experiment using a tactile-equipped robotic arm and a Long Short-Term Memory (LSTM) model to detect object presence.
Drawing from granular media particle interaction theory, we hypothesize tactile cues extend up to 7 cm.
Human results confirm detection with 70.7% precision at a 6.9 cm distance (median 2.7 cm). Robotic results align, with the LSTM model detecting objects at 7.1 cm, though with a better median (6 cm) but lower precision (40%).
This work enhances understanding of human tactile perception in granular media and introduces a robotic system to study human exploration strategies and enable autonomous applications in archaeology, space exploration, and search and rescue.