Electric fish in augmented reality environment using active sensing.

The Electric Sense: How Fish Use Movement to 'See' Their World

"Electric fish show us that active sensing, blending movement and perception, is key to understanding how we all interact with our environment"


Our senses constantly feed us information about the world. What’s often overlooked is how much our own movements shape this sensory input. Sometimes, movements stabilize what we perceive, like when our eyes compensate while tracking a moving object. Other times, movements actively enhance the information we gather. Consider when you're searching for keys in your pocket – the complex motions of your hand and fingers help you discern shape, texture, and weight.

This 'active sensing' is widespread across different senses and contexts. However, how we control and adjust these movements based on the situation – say, feeling for a dull versus a sharp object – remains a puzzle. Fortunately, a recent study featured in Current Biology offers fascinating clues. Biswas et al. [3] explored weakly electric fish in an augmented reality setting, revealing how these animals dynamically adjust their movements in response to the sensory feedback they receive.

Gymnotiform weakly electric fish, such as Eigenmannia virescens, use their active electric sense to explore. They generate an electrical field around their body and detect distortions caused by objects in their environment using electroreceptors on their skin. These fish often hide in plant thickets during the day to avoid predators, tracking movements in the thickets to maintain a stable position [4]. The group of Eric Fortune and Noah Cowan has examined how these fish track moving refuges in controlled lab settings [5].

AI Search Multiple angles on this topic

Active Sensing as a Concept

Active sensing refers to the process of actively scanning the environment with one's senses rather than passively waiting for stimuli. This approach is used across multiple domains, from engineering systems to biological organisms. Research into active sensing explores how agents choose the most informative samples and viewpoints to acquire information about their surroundings. The concept spans both technological applications and fundamental questions about perception.

Methodological Challenges in Sensing Systems

Current sensing approaches face significant methodological limitations, including challenges with standardization and harmonization across platforms. Method heterogeneity and analytical differences across platforms can confound direct comparisons and limit generalization. Active sensing methods must contend with dimension limitations in detectability, particularly when dealing with complex multi-layered systems. These challenges affect both biological sensing research and technological sensing applications.

Origins and Development

The concept of milestones has deep historical roots, with the word deriving from Old English and originally referring to physical stone markers used for measuring distance. In scientific and technological contexts, milestones represent key achievements that mark progress in a field of study. The development of sensing technologies has followed its own trajectory of milestones, building on foundational discoveries in physics and biology. Understanding these origins provides context for current advances in active sensing research.

The Active Sensing Dance: Movement and Perception

Electric fish in augmented reality environment using active sensing.

The researchers discovered that these fish perform distinct longitudinal movements, almost like a 'va-et-vient', in addition to tracking the refuge [6]. These movements seem counterintuitive, as they require extra energy and appear to destabilize sensory input. So why do the fish do it? Further investigation showed that these movements, which alter the relative motion between the fish and refuge, are carefully controlled. The amplitude of these movements decreases when visual information is available and increases when water conductivity drops (reducing the effectiveness of electrosensory input). This suggests the movements are part of an active sensing strategy.

Studying active sensing is tricky because the animal's behavior directly influences the sensory input it receives. This calls for clever experimental designs that allow the animal to behave naturally while also enabling precise manipulation of sensory feedback. Biswas et al. [3] tackled this challenge by creating an augmented reality system. This system allowed them to manipulate the sensory feedback the fish received based on its own movements [3]. Specifically, they linked the refuge's motion to the fish's motion and systematically changed the gain of this interaction. For instance, a negative gain made the refuge move in the opposite direction to the fish, while a positive gain caused it to move in the same direction.

  • Increased Sensing Volume: Movements can expand the area in which fish can detect stimuli.
  • Counteracting Adaptation: Movements help overcome the sensory system's tendency to filter out constant stimuli.
  • Information Creation: Active movements may generate new types of information about the surrounding environment.
AI Search Multiple angles on this topic

Current Research Frontiers

Active sensing research continues to advance across multiple disciplines, with recent studies exploring optimal array geometries and waveform designs to minimize error bounds. Research at institutions like the Conte Center for Active Sensing focuses on integrative approaches to understanding how organisms actively scan their environments. Scientists investigate how rather than passively receiving information, agents use movement and sensory organs to look around and gather data. This research spans both biological systems and engineered sensing technologies.

Limitations and Critiques

Active sensing systems face several critical limitations that researchers must acknowledge. Current AI agents have fundamental constraints in their ability to process and respond to sensory information in real-time. These limitations affect both technological implementations and our understanding of biological sensing systems. Researchers continue to identify failure modes and boundary conditions where active sensing approaches may not perform optimally.

Comparing Sensing Approaches

Different sensing approaches offer distinct advantages and tradeoffs depending on the application. Active monitoring systems differ from passive approaches in how they engage with the environment and gather data. Animal senses vary significantly across species, with each adapted to specific ecological niches and survival requirements. These comparisons help researchers understand the relative strengths of different sensing strategies in both natural and artificial systems.

The most striking finding was that the fish dramatically altered the amplitude of their own movements to keep the variance in sensory input (sensory slip) consistent, regardless of the augmented reality gain (Figure 2B). These results strengthen the idea that sensory slip isn't just an error in tracking but a deliberate outcome of active sensing. More importantly, it suggests that the fish extract information from this self-generated sensory slip to aid in their tracking. One idea is that these movements increase the sensing volume, or the area within which the fish can detect stimuli [7]. Another possibility is that these movements counteract the adaptation properties of the fish's electrosensory neurons [8]. These neurons are more responsive to changing signals but tend to filter out constant signals [9], like the steady movement of a refuge. The active movements might reintroduce these signals at higher frequencies, making them easier to detect. The most intriguing possibility is that these movements create unique information about the environment that wouldn't be available otherwise, which the electrosensory system then decodes [10].

The Future of Active Sensing Research

To understand how active sensing movements enhance sensory input, future research needs to record neural activity in freely swimming fish during refuge tracking. Recent technological advances have made such recordings increasingly feasible in aquatic animals [11]. Sensory input is a mix of external (ex-afferent) and self-generated (re-afferent) signals, and a key question is how the nervous system distinguishes between them during processing. In refuge tracking, the sensory input from the refuge's movement is ex-afferent, while the input from the 'va-et-vient' movements is re-afferent.

AI Search Multiple angles on this topic

Expert Perspectives

Remote sensing expertise requires skills in data processing including layer stacking, mosaicking, and ortho-rectification of satellite imagery. Experts apply GIS and remote sensing techniques for interpretation and visual digitalization of very-high-resolution data. The field demands quality assurance processes and systematic approaches to data analysis. These professional perspectives highlight the technical rigor required in modern sensing applications.

Emerging Trends

The touchless sensing market is evolving through integration with AI and IoT ecosystems, with increased focus on user-centric design. Remote sensing of night lights offers unique opportunities to observe human activity from space, enabling applications from urban mapping to disaster monitoring. Energy sector sensing capabilities continue advancing with growth in renewable energy monitoring. These trends suggest a future where sensing technologies become more integrated and ubiquitous.

Systemic Considerations

Urban sensing systems present choices between opportunistic and participatory approaches, each with different requirements for user engagement. Participatory sensing relies on active user involvement in data collection decisions, while opportunistic sensing reduces these demands. These approaches operate within broader systemic contexts where leadership decisions cascade through interconnected systems. Understanding these dynamics is crucial for designing effective sensing frameworks.

Real-World Applications

Augmented reality technology has been used to study how fish use movement-based active sensing to perceive their environment. Research demonstrated that nearly all animals use movement to actively sense the world around them, representing a fundamental process of perception. This approach allows researchers to probe in real time how organisms interact with their surroundings through active sensing. The study represents perhaps the first use of augmented reality for investigating this fundamental biological process.

About this Article -

Written with AI assistance from published research, and reviewed by the Mystum team. See our About page for more information.

This article is based on research published under:

DOI-LINK: 10.1016/j.cub.2018.10.060, Alternate LINK

Title: Active Sensing: Constancy Requires Change

Subject: General Agricultural and Biological Sciences

Journal: Current Biology

Publisher: Elsevier BV

Authors: Volker Hofmann, Maurice J. Chacron

Published: 2018-12-01

Everything You Need To Know

1

How do electric fish utilize active sensing to explore their environment?

Electric fish like *Eigenmannia virescens* use active sensing by generating an electrical field around their bodies. They possess electroreceptors that detect distortions in this field caused by objects. This active electric sense helps them to navigate, locate objects, and maintain a stable position, particularly when hiding in plant thickets to avoid predators. This combination of generating an electric field and sensing distortions is central to how these fish perceive their surroundings.

2

What distinct movements do electric fish perform while tracking a refuge, and how are these movements controlled?

Researchers discovered that electric fish perform longitudinal movements, resembling a 'va-et-vient,' while tracking a refuge. These movements, controlled by the fish, alter the relative motion between the fish and the refuge. The amplitude of these movements decreases when visual information is available and increases when water conductivity drops, which reduces the effectiveness of electrosensory input. This indicates that these movements are part of a carefully managed active sensing strategy.

3

What is 'sensory slip' in the context of electric fish, and why is it considered important for active sensing?

Sensory slip, in the context of electric fish, refers to the variance in sensory input that arises from the fish's own movements. The research indicates that sensory slip isn't just an error in tracking, but is instead a deliberate outcome of active sensing. The fish appear to extract information from this self-generated sensory slip to aid in tracking, which helps them understand changes in their environment and potentially enhances their ability to detect stimuli.

4

How did the study by Biswas et al. [3] use augmented reality to investigate the active sensing of electric fish?

The study by Biswas et al. [3] used an augmented reality system to investigate the active sensing of electric fish. This system allowed researchers to manipulate the sensory feedback the fish received based on its own movements, specifically by linking the refuge's motion to the fish's motion and systematically changing the gain of this interaction. For example, they could make the refuge move in the opposite direction to the fish (negative gain) or in the same direction (positive gain). This manipulation enabled precise control over the sensory input the fish received during tracking.

5

What are the key areas of focus for future research on active sensing in electric fish, and what challenges need to be addressed?

Future research should focus on recording neural activity in freely swimming fish during refuge tracking to understand how active sensing movements enhance sensory input. Because sensory input is a mix of external (ex-afferent) and self-generated (re-afferent) signals, it is important to determine how the nervous system distinguishes between these signals during processing. The sensory input from the refuge's movement is ex-afferent, while the input from the 'va-et-vient' movements is re-afferent. Advances in technology are making such recordings increasingly feasible in aquatic animals.

Newsletter Subscribe

Subscribe to get the latest articles and insights directly in your inbox.