Quadrupedal animals like mice navigate their environments through complex coordination of neural signals and biomechanical movements, enabling stable and directed locomotion. While many computational models simplify this process by assuming left-right symmetrical body movements and focusing on straight-line paths, real animals rely heavily on asymmetrical body movements to execute turns and adjust speed effectively. This study builds upon a previously developed model of quadrupedal locomotion proposed by Molkov et al., 2024) in which forward movement of the body was driven by central neural interactions, biomechanics, and proprioceptive feedback. We extended this model to comparatively investigate possible mechanisms of steering by introducing three distinct asymmetrical strategies–body bending, lateral force application, and lateral limb shifting as well as their combinations–to explore their potential involvement in turning performance. By simulating these strategies across a walking speed range, we measured and compared their impact on turning curvature the sharpness of the turn) and limb coordination. The latter was quantified through ratios of duty factors representing the relative time that a limb spent in contact with the ground compared to its counterpart on the opposite side. Our findings reveal that each strategy excels at different speeds: body bending allows sharp turns at low speeds, lateral force is most effective at medium speeds, and lateral shifting performs best at higher speeds. Our results suggest that animals select or combine turning strategies based on their locomotor speed or adjust speed to use a specific strategy. We also show that the forelimbs consistently play a primary role in steering, while the hindlimbs adjust propulsion and stability in ways that depend on the specific turning strategy. These results provide valuable insights into how spinal circuits and mechanical asymmetries work together to produce flexible, adaptive movement patterns, offering a robust framework for understanding locomotion in both biological organisms and robotic systems designed to mimic such behaviors.

Animal intelligence is not purely a product of abstract computation in the brain, but emerges from dynamic interactions between the nervous system and the body. New connectome datasets and musculoskeletal models now enable integrated, closed-loop simulations of the neural and biomechanical systems of the fruit flyDrosophila, an ideal model organism to investigate embodied intelligence. However, many biological parameters of the nervous system and the body, as well as how they interface, remain unknown. To fill such gaps, researchers are turning to deep reinforcement learning (DRL), a data-driven optimization framework, to create virtual animals that imitate the behavior of real animals. Here, we provide a cautionary tale about the interpretation of such models. We constructed a virtual chimera of two phylogenetically distant species: a connectome of theC. elegansnematode worm and a biomechanical model of the fly body. The worm connectome receives sensory information from the fly body, and an artificial neural network is trained with DRL to map worm motor neuron activations to the fly’s leg actuators. The resulting digital sphinx produces highly realistic fly walking—yet it is biologically meaningless. This exercise teaches us nothing about either animal and exposes a core peril of connectome-body models: behavioral fidelity is achievable without biological fidelity, making such models easy to overinterpret. Done carefully, virtual animals can be powerful partners to biological experiments, but only if their components and interfaces are grounded in biology.

In typical models of collective motion, each individual takes the average direction of multiple neighbors, resulting in the ordered movement of large flocks. Alternatively, interactions with only one random neighbor at a time can also lead to order, referred to as noise-induced order, but only in small flocks. Here, we propose a new model of collective movement in which individuals interact with a single random neighbor yet can produce order in large flocks. Our model consists of individuals who can also stop, inspired by the intermittent stop-and-go motion in real animals. We introduce a halting interaction in which an individual may stop upon encountering a neighbor moving in the opposite direction, rather than instantly aligning. Using mean-field deterministic and stochastic approximations for a one-dimensional version of the model, we show that persistent collective order emerges even in large flocks. This represents a different mechanism from conventional averaging-based order in large flocks or noise-induced ordering in small flocks. We confirm the mean-field predictions using individual-based simulations in both one and two dimensions. Our results highlight how incorporating a stopped state and halting interactions can generate new routes to order in collective movement.

This work proposes an intelligent and low-cost solution to detect and respond to wild animal intrusions, particularly by elephants, in agricultural regions, addressing significant threats to crops, livelihoods, and human safety. The system utilizes a ground-based detection unit featuring an ESP32-CAM module equipped with Infrared (IR) night vision to continuously monitor the field day and night. Upon detecting an animal, the system sends location information to an autonomous drone which navigates to the spot. The drone is equipped with bright flashing lights and a sound device for effective, non-lethal repelling. Simultaneously, a GSM module (SIM800L) makes an automated phone call alert to the farmer, ensuring immediate awareness even on keypad mobile phones. This affordable, efficient, and reliable IoT-based solution operates with minimal human intervention, aiming to significantly reduce crop damage and increase safety in rural farming communities through real time animal detection and response.

Understanding the depth of penetration of near-infrared (NIR) light in biological tissue is critical for enhancing clinical applications of near-infrared spectroscopy (NIRS). The current knowledge of NIRS penetration depth primarily stems from mathematical models, numerical simulations, and phantom studies, with a notable knowledge gap derived from real animal models. By sequentially obstructing light from traversing in a porcine kidney tissue model, we derived the depth distribution of NIR light experimentally and better characterized its dependence on the distance between the light source and photodetector. We collected four replicates of data from six different source-detector distances (SDSs) and found that both the maximum and mean depths of penetration of NIRS increase with the SDS. Linear relationships can be derived between the SDS and the maximum depth, and the square root of the SDS and the mean depth.

This article critically analyzes Ecuador's legal framework on animal abuse and killing, reviewing the 2008 Constitution, the COIP (2014/2019), the CODA (2017), and landmark cases such as “Mona Estrellita” and “Spayk.” It also compares Ecuador's legislation with that of Spain, Colombia, the United Kingdom, and the Netherlands. It highlights procedural gaps, such as private criminal action, and institutional shortcomings that limit the effectiveness of sanctions and administrative protection. Although regulatory advances and penalties of up to three years for cruel death are recognized, impunity and inequality in territorial application persist. Comprehensive reforms are proposed: declaring abuse crimes public, increasing penalties, creating an Organic Law on Animal Welfare, establishing specialized units, registering animals and abusers, and strengthening education and shelters. These measures seek to translate the legal framework into real and effective animal protection.

Based on in-depth interviews with care workers and observational visits to nursing homes, this study investigates how care workers address residents’ frequent misperceptions of robot cats and dogs as real animals. The analysis focuses on two aspects: how care workers handle the fact that residents often mistake the robots for real animals, and how their approach to deceptive practices relation to the robots is related to emotional labor. Three main strategies are identified and explored: telling the truth, remaining vague, and lying. While the first strategy prioritizes ethical guidelines over residents’ wellbeing in the moment, the second two strategies are facilitated by physical and verbal cues, as well as storytelling in collaboration with colleagues and residents. Each strategy also entails a dilemma, as each carries its own ethical challenges.

Simple SummaryMonitoring animal welfare is a key element of conducting research involving animals. The aim of this project was to co-create animal welfare monitoring systems that could contribute to best practice husbandry standards of farm animals in a real animal research setting. Researchers worked with nine staff to co-design six bespoke welfare assessment protocols to be conducted in addition to legally required welfare monitoring for adult cattle, calves, sheep, pigs, and goats in specific experimental environments that included both positive and negative welfare elements. Four protocols were subsequently applied with variable frequency by three staff to cattle, goats, and two pig populations. Assessments were all observational and included behavioural and physical condition data. Two staff provided feedback on their views of the process. A key finding was that with facilitation, staff could generate protocols that included elements designed to encourage or evaluate interventions to promote positive emotions. However, data collection was sporadic, and although the staff who provided feedback reported that they valued the process highly, they noted that the primary challenge was finding the time to conduct the additional assessments. We therefore conclude that sustained support is likely to be required to maximise the benefits for the animals and staff of developing and conducting additional voluntary welfare monitoring of farm animals.The aim of this project was to co-create an animal welfare monitoring system that incorporated both positive and negative welfare measures that would contribute to best practice husbandry standards of farm animals in a real animal research setting. Researchers worked with nine staff to co-design six bespoke welfare assessment protocols to be conducted in addition to legally required welfare monitoring for adult cattle, calves, sheep, pigs, and goats in specific experimental environments. Four protocols were subsequently applied with variable frequency by three staff to cattle, goats, and two pig populations. Assessments were all observational, and included behavioural scan sampling, Qualitative Behaviour Assessment scores, visual analogue mood scores, and physical condition data. Two staff provided feedback on their views of the process. A key finding was that with facilitation, staff could generate protocols that included elements designed to encourage or evaluate interventions to promote positive emotions. However, data collection was sporadic, and although the staff who provided feedback reported that they valued the process highly, they noted that the primary challenge was finding the time to conduct the assessments. We therefore conclude that sustained support is likely to be required to maximise the benefits for the animals and staff of developing and conducting voluntary welfare monitoring of farm animals.

Iterative optimization algorithm with structural prior for artifacts removal of photoacoustic imaging.

Developmental differences in the timecourse of word learning: Greater improvements for children, semantic benefits for adults.

This paper presents the proposed prototype of an Animal Intrusion Detection System powered by Artificial Intelligence of Things (AIoT) technology to address growing challenges of human-wildlife conflicts (HWC) in Bhutan. The major incursions of wildlife in the agriculture fields possess a major threat to sustainable food security and farmer livelihoods in the country. While the government has implemented various mitigation measures like electric and chain-link fencing, and animal repellent system, these solutions have notable limitations. Therefore, our AI-based system aims to provide as an alternative smart agriTech solution to address HWC. The system utilizes a Raspberry Pi 4, a night vision-based camera, an ultrasonic sensor and YOLOv8 deep learning algorithm for real-time animal detection and classification. The YOLO model was trained on a dataset of 30,800 images featuring seven local wildlife species which are common in raiding the crop in Bhutan. The system, upon detecting an intrusion on farmland, will automatically transmits an alert notification to farmers via a mobile app over a cellular network, enabling timely intervention to mitigate the crop damage. When the internet connection is down, the system will notify the farmers through SMS and Dial. In a controlled laboratory environment, the prototype achieved a detection accuracy of 95.7%. These finding indicates a promising alternative innovative agriTech solution for mitigating crop losses, enhancing food security and enhancing farmer livelihoods. However, the prototype requires field validation and further AI model training with a more extensive real animal dataset collected through its pilot implementation to evaluate the system's performance and robustness under real-world conditions of the agriculture field.

Alginate-gelatin hydrogel scaffolds for establishing physiological barriers on a gut-brain-axis microchip.

The AI-Based Scarecrow project is an innovative solution designed to address the critical issue of crop damage caused by wildlife. By leveraging real time animal detection and automated deterrence systems, this project provides an effective, efficient, and scalable method to protect crops from animals such as deer, monkeys, elephants, and other species that pose a threat to agricultural fields. At the core of the system is the YOLOv3 deep learning algorithm, a state-of-the-art object detection model known for its accuracy and speed. The algorithm processes live video feeds captured by cameras deployed in the fields, identifying intruding animals in real time. Once an animal is detected, the system triggers species specific deterrent sounds designed to scare the animal away, thereby preventing it from causing damage to the crops. These sounds include natural predator calls or other noise stimuli that are known to be effective in repelling different types of wildlife. This approach reduces the reliance on traditional methods, such as electric fencing or manual surveillance, which are labor-intensive, costly, and often less effective.

Machine learning has attracted much attention in the field of genomic prediction due to its powerful predictive capabilities, yet the lack of an explanatory nature in modeling decisions remains a major challenge. In this study, we propose a novel machine learning method, ExAutoGP, which aims to improve the accuracy of genomic prediction and enhance the transparency of the model by combining automated machine learning (AutoML) with SHapley Additive exPlanations (SHAP). To evaluate ExAutoGP’s effectiveness, we designed a comparative experiment consisting of a simulated dataset and two real animal datasets. For each dataset, we applied ExAutoGP and five baseline models—Genomic Best Linear Unbiased Prediction (GBLUP), BayesB, Support Vector Regression (SVR), Kernel Ridge Regression (KRR), and Random Forest (RF). All models were trained and evaluated using five repeated five-fold cross-validation, and their performance was assessed based on both predictive accuracy and computational efficiency. The results show that ExAutoGP exhibits robust and excellent prediction performance on all datasets. In addition, the SHAP method not only effectively reveals the decision-making process of ExAutoGP and enhances its interpretability, but also identifies genetic markers closely related to the traits. This study demonstrates the strong potential of AutoML in genomic prediction, while the introduction of SHAP provides actionable biological insights. The synergy of high prediction accuracy and interpretability offers new perspectives for optimizing genomic selection strategies in livestock and poultry breeding.

Abstract Companion animals significantly enhance human mental health, which became especially evident during the COVID -19 pandemic. Although there has been a global surge in online animal-watching in recent years, little research has been conducted to explore its impact. This study, grounded in attachment theory, aimed to investigate the effects of keeping real and online (“cloud”) animals on the mental health of Chinese adults. The results revealed that real and cloud animal keepers experienced more attachment to their animals than non-keepers, and greater attachment was associated with better mental health. However, real animal keepers had significantly better mental health than non-keepers and cloud animal keepers. For real animal keepers, greater social support played a positive role in moderating the relationship between attachment to animals and mental wellbeing. However, this moderating effect was not significant for those engaged in online animal-watching. The findings should encourage research on potential influences of cloud animal keeping.

Georges Horan-Koiransky, in Journal d’un interné. Drancy 1942-1943, makes a recur-rent use of images of real and fantastical animals to evoke his imprisonment. Two parallel and complementary perspectives, psychoanalysis and mythocriticism, shall provide the means for re-vealing and analysing the issues at stake. Psychoanalysis will enable us to establish a link between the use of figurative language and the post-traumatic stress disorders resulting from Georges’ imprisonment in the Police aux Questions Juives (PQJ) and the transit camps. Once we have examined the psyche of the diarist, a mitocritical approach based on Gilbert Durand’s suggestions on the imagination will enable us to reveal the presence of zoomorphic images in his apprehen-sion and interpretation of the universe within the walls of the places of imprisonment.

ABSTRACTPopular in fantasy films, games, and extended reality, anthropomorphic animals often rely on animator creativity and real animal observation for behavior visualization. This artistic approach captures emotional traits but lacks uncovering diverse, unanticipated behaviors beyond creators' concepts. To enrich character design, this study employs reinforcement learning (RL) agent simulation to explore the autonomous behavior and unexpected responses of the nine‐tailed Fox Sister from Korean folklore. As a method, the agent, with a physics‐based controller and skeletal joints, uses hybrid action control to transition between bipedal and quadrupedal actions based on the environment. In result, RL character frequently exhibits behavioral shifts, including unexpected actions in response to training steps and terrain complexities like slopes and hurdles, distinguishing them from animation‐based finite‐state machines. Additionally, this study validates impacts of RL character on character design creativity. To investigate such unknown impacts, this study conducts a comparative pilot study that recruits five character designers under use and nonuse scenario of RL character. Analysis indicates that RL character promotes creativity of character design, conceptualization, and development of scenario and character's attribute. This study highlights RL's potential for visualizing diverse inspirational behaviors of folkloric creatures by simulating interactions between body structure, motion, and environment.

Abstract Taxidermy, first practised in the sixteenth century, was often used to preserve animal remains in museums of the European Renaissance, and was soon deployed to create dragon hoaxes. Many such specimens were displayed in museums and were cited by naturalists as evidence that dragons were real animals. The dracunculus of King Francis I of France (r. 1515–47) was an internationally famous case. The specimen is now lost, but a previous investigation of contemporary drawings and a written description showed that the specimen had the skull of a weasel and the tail skeleton of an eel. This continuation of the investigation reveals that the limbs, skin and ribcage were those of an ocellated lizard (Timon lepidus), with the skin shifted over the lizard’s skeleton into an unnatural position. This study elucidates the materials and methods that were used to produce an astonishing category of biological fraud in which early museums unwittingly participated.

The continual generation of behaviors that satisfy all conflicting demands that cannot be satisfied simultaneously, is a situation that is seen naturally in autonomous agents such as long-term operating household robots, and in animals in the natural world. Homeostatic reinforcement learning (homeostatic RL) is known as a bio-inspired framework that achieves such multiobjective control through behavioral optimization. Homeostatic RL achieves autonomous behavior optimization using only internal body information in complex environmental systems, including continuous motor control. However, it is still unknown whether the resulting behaviors actually have the similar long-term properties as real animals. To clarify this issue, this study focuses on the balancing of multiple nutrients in animal foraging as a situation in which such multiobjective control is achieved in animals in the natural world. We then focus on the nutritional geometry framework, which can quantitatively handle the long-term characteristics of foraging strategies for multiple nutrients in nutritional biology, and construct a similar verification environment to show experimentally that homeostatic RL agents exhibit long-term foraging characteristics seen in animals in nature. Furthermore, numerical simulation results show that the long-term foraging characteristics of the agent can be controlled by changing the weighting for the agent’s multiobjective motivation. These results show that the long-term behavioral characteristics of homeostatic RL agents that perform behavioral emergence at the motor control level can be predicted and designed based on the internal dynamics of the body and the weighting of motivation, which change in real time.

With the advancement in relative technology, robotics is gaining an increasing amount of attention among engineers and researchers as it is a current focus on enabling peoples lives to become more convenient. However, there is another type of robot designed very differently from functional robots like delivery robots and quadrotors that most people are familiar with, bionic robots. This is a kind of robots that gets design inspiration from real animals in nature, from the design of structure to altitudes and mechanisms of motion. How bionic robots are designed, and how the conceptional difference of bionic robots with other robots is reflected on their physical designs are the focuses of this paper. In this paper, four basic types of bionic robots are introduced and 17 existing designs of bio-inspired vehicles are included. 10 bionic aerial vehicles in 3 smaller classifications are compared to 3 conventional aerial vehicles in order to figure out differences in 3 aspects (structure design, sensors, and control method). The results show that bionic robots usually have more degrees of freedom in order to mimic animals attitudes, and are lighter in weight. Typically, bio-inspired aerial vehicles employ different wing systems compared with conventional ones. Additionally, while conventional vehicles favour PID more because of its ease of construction, bionic robots use machine learning techniques. Both types of robots use the same sensors, which is necessary for them to detect their environment. However, their control methods differ due to different purposes.