Robot Behavior Research Articles

Printing Untethered Self-Reconfigurable, Self-Amputating Soft Robots from Recyclable Self-Healing Fibers.

Regarding the challenge of self-reconfiguration and self-amputation of soft robots, existing studies mainly focus on modular soft robots and connection methods between modules. Different from these studies, this study focus on the behavior of individual soft robots from a material perspective. Here, a kind of soft fibers, which consist of hot melt adhesive particles, magnetizable microparticles, and ferroferric oxide microparticles embedded in a thermoplastic polyurethane matrix are proposed. The soft fibers can achieve wireless self-healing and reversible bonding of the fibers by eddy current heating and can be actuated by magnetic fields. Moreover, the soft fibers are recyclable and printable. Building on this material foundation, an integrated material-structure-actuation printing strategy using soft fibers for the design and fabrication of soft robots are reported. The robots printed by this strategy can achieve their untethered motions and wireless self-healing. Soft gripper, soft crawling robot, and soft multi-legged robot, are then fabricated which demonstrates the self-healing, self-reconfigurable, self-amputating, and sustainable performances of soft robots so as to adapt to different environments and tasks. This integrated material-structure-actuation printing strategy using soft fibers is universal, easy to implement, and mass-manufactured, opening a door for sustainable, eco-friendly, untethered, self-reconfigurable, self-amputating soft robots.

Personalization of Robot Behavior Using Approach Based on Model Predictive Control

This paper proposes a novel approach to personalizing robot behavior using Model Predictive Control (MPC). Social humanoid robots, equipped with advanced sensors and human-like capabilities, are increasingly integrated into human environments, necessitating adaptable and intuitive communication interfaces. Our approach enables the design of adaptive interfaces that support fluid, personalized human–robot interactions. In the proposed framework, a user model is applied to predict responses to potential robot actions. Initially, this model represents an average user; however, it is updated as the robot gathers new observations, leading to increasingly personalized decisions. Experiments assessed the performance of five machine-learning algorithms generating user models in a simulated environment, with the Light Gradient Boosting Machine (LGBM) achieving the best results, closely followed by Random Forest (RF). A comparison of the inference time showed that LGBM is more than four times faster than RF. Outlier-removal techniques showed a modest performance improvement over models without outlier removal. Additionally, robot adaptation was tested in the experiments, showing an increase in the average reward over time, although with a relatively high standard deviation. The results suggest that the proposed approach for robot behavior adaptation based on MPC works well, and the recommended algorithm for the user model is LGBM.

Meet the motivational robot that predicts your future feelings

This study explores the potential benefits of robots having the capability to anticipate people’s mental states in an exercise context. We designed 80 utterances for a robot with associated gestures that exhibit a range of emotional characteristics and then performed a 23-person data collection to investigate the effects of these robot behaviors on human mental states during exercise. The results of cluster analysis revealed that 1) utterances with similar meanings had the same effect and 2) the effects of a certain cluster on different people depend on their emotional state. On the basis of these findings, we proposed a robotic system that anticipates the effect of utterances on the individual’s future mental state, thereby choosing utterances that can positively impact the individual. This system incorporates three main features: 1) associating the relevant events detected by sensors with a user’s emotional state; 2) anticipating the effects of robot behavior on the user’s future mental state to choose the next behavior that maximizes the anticipated gain; and 3) determining appropriate times to provide coaching feedback, using predefined rules in the motion module for timing decisions. To evaluate the proposed system’s overall performance comprehensively, we compare robots equipped with the system’s unique features to those lacking these features. We design the baseline condition that lacks these unique features, opting for periodic random selection of utterances for interaction based on the current context. We conducted a 21-person experiment to evaluate the system’s performance. We found that participants perceived the robot to have a good understanding of their mental states and that they enjoyed the exercises more and put in more effort due to the robot’s encouragement.

Tactile Interaction with Social Robots Influences Attitudes and Behaviour

Tactile Interaction with Social Robots Influences Attitudes and Behaviour

Time-Varying Preference Bandits for Robot Behavior Personalization

Robots are increasingly employed in diverse services, from room cleaning to coffee preparation, necessitating an accurate understanding of user preferences. Traditional preference-based learning allows robots to learn these preferences through iterative queries about desired behaviors. However, these methods typically assume static human preferences. In this paper, we challenge this static assumption by considering the dynamic nature of human preferences and introduce the discounted preference bandit method to manage these changes. This algorithm adapts to evolving human preferences and supports seamless human–robot interaction through effective query selection. Our approach outperforms existing methods in time-varying scenarios across three key performance metrics.

Swarm Intelligence in Action: Particle Swarm Optimization and Rendezvous Algorithms for Swarm Robotics

ABSTRACTSwarm technology is evolving quickly in the new era in the domains of autonomous underwater vehicles (AUVs), unmanned aerial vehicles (UAVs), and unmanned ground vehicles (UGVs). Swarm technology takes its cues from nature, particularly from the behaviors of ants, bees, schools of fish, and birds. The rapid growth of swarm technologies can be attributed to this organic inspiration. It is utilized in a variety of applications, such as pick‐and‐place robotics and drone swarms for laser shows. To mimic the behavior of swarm robots or microbots, this research work explores the use of swarm algorithms, namely particle swarm optimization (PSO) and rendezvous algorithms (RA). The emphasis will be on applying these algorithms to situations that require swarm robots to navigate around obstacles while planning their paths and forming patterns. This study shows how successful these algorithms are by using MATLAB as a useful tool. The outcomes demonstrate that the algorithms are capable of producing intricate patterns and facilitating effective path planning for swarms of robots. The robots' ability to avoid obstacles and travel around them at a high speed is noteworthy. This study covers a wide range of industries, such as space exploration, military/defense applications, storage, surveillance, and search and rescue. The application of swarm technology demonstrates how it may transform practical situations and provides information about how flexible and efficient swarm algorithms are in a variety of settings. By offering useful insights for researchers, engineers, and practitioners in utilizing the combined potential of swarm intelligence (SI) for a variety of applications, this investigation advances the field of swarm robotics.

A Framework for Simultaneous Task Allocation and Planning under Uncertainty

We present novel techniques for simultaneous task allocation and planning in multi-robot systems operating under uncertainty. By performing task allocation and planning simultaneously, allocations are informed by individual robot behaviour, creating more efficient team behaviour. We go beyond existing work by planning for task reallocation across the team given a model of partial task satisfaction under potential robot failures and uncertain action outcomes. We model the problem using Markov decision processes, with tasks encoded in co-safe linear temporal logic, and optimise for the expected number of tasks completed by the team. To avoid the inherent complexity of joint models, we propose an alternative model that simultaneously considers task allocation and planning, but in a sequential fashion. We then build a joint policy from the sequential policy obtained from our model, thus allowing for concurrent policy execution. Furthermore, to enable adaptation in the case of robot failures, we consider replanning from failure states and propose an approach to preemptively replan in an anytime fashion, replanning for more probable failure states first. Our method also allows us to quantify the performance of the team by providing an analysis of properties, such as the expected number of completed tasks under concurrent policy execution. We implement and extensively evaluate our approach on a range of scenarios. We compare its performance to a state-of-the-art baseline in decoupled task allocation and planning: sequential single-item auctions. Our approach outperforms the baseline in terms of computation time and the number of times replanning is required on robot failure.

Influence of the Behavior of a Para-Operated Social Robot on the Impression Interlocutors Have of the Operator: an Online Study

Influence of the Behavior of a Para-Operated Social Robot on the Impression Interlocutors Have of the Operator: an Online Study

EchoPT: A Pretrained Transformer Architecture That Predicts 2D In-Air Sonar Images for Mobile Robotics.

The predictive brain hypothesis suggests that perception can be interpreted as the process of minimizing the error between predicted perception tokens generated via an internal world model and actual sensory input tokens. When implementing working examples of this hypothesis in the context of in-air sonar, significant difficulties arise due to the sparse nature of the reflection model that governs ultrasonic sensing. Despite these challenges, creating consistent world models using sonar data is crucial for implementing predictive processing of ultrasound data in robotics. In an effort to enable robust robot behavior using ultrasound as the sole exteroceptive sensor modality, this paper introduces EchoPT (Echo-Predicting Pretrained Transformer), a pretrained transformer architecture designed to predict 2D sonar images from previous sensory data and robot ego-motion information. We detail the transformer architecture that drives EchoPT and compare the performance of our model to several state-of-the-art techniques. In addition to presenting and evaluating our EchoPT model, we demonstrate the effectiveness of this predictive perception approach in two robotic tasks.

A modified extended Kalman filter based fusion of auto-correlation data for robot SLAM estimation

Abstract. With the popularization of robot technology, various technological developments related to robots have gradually entered a stage of rapid development. Especially path planning technology, which has been a very important and worthwhile part of robot motion and behavior since the concept of robots emerged. In this technology, various obstacles, environmental changes, and efficiency factors usually need to be considered. Whether in the past, present, or future, robot behavior pursues greater precision, reliability, adaptability to the environment, and even anthropomorphism. In order to achieve these goals, the idea of data fusion has emerged in the development of robot data processing. Therefore, the research direction is to use Kalman filtering for noise filtering and data correction to improve the accuracy of fused data. By using a feasible and strongly correlated data fusion method, combined with an improved extended Kalman filter, various data generated by robots can be fused in multiple dimensions and have correlations, reducing the negative impact of data differences and improving the quality and reliability of data fusion.

The Journey or the Destination: The Impact of Transparency and Goal Attainment on Trust in Human-Robot Teams

As robots gain autonomy, human-robot task delegation can become more goal-oriented; specifying what to do rather than how. This can lead to unexpected robot behaviour. We investigated the effect of transparency and outcome on the perceived trustworthiness of a robot that deviates from the expected manner to reach a delegated goal. Participants (N = 82) engaged in a virtual military mission as a Human-Robot Team using a 2x2 between-subjects design (low vs. high transparency, positive vs. negative outcome). Participants received training on the expected manner to reach the mission’s goal. In the actual mission, the robot deviated from the planned path. We manipulated whether the robot explained its deviation and whether the outcome was better or worse than the original plan. Results showed that transparency contributed to higher and more stable levels of trust, without increasing subjective workload. While the robot’s deviation led to a violation of trust in the low transparency condition, trust remained stable in the high transparency condition, indicating a buffering effect of transparency on trust in case of unexpected behaviour. The impact of outcome on trust was consistent across transparency conditions. Our findings underscore the role of transparency as a tool for fostering human-robot trust.

An empirical evaluation of a formal approach versus ad hoc implementations in robot behavior planning

As autonomous robotic systems integrate into various domains, ensuring their safe operation becomes increasingly crucial. A key challenge is guaranteeing safe decision making for cyber-physical systems, given the inherent complexity and uncertainty of real-world environments.Tools like Gwendolen, vGOAL, and Tumato enable the use of formal methods to provide guarantees for correct and safe decision making. This paper concerns Tumato, a formal planning framework that generates complete behavior from a declarative specification. Tumato ensures safety by avoiding unsafe actions and states while achieving robustness by considering nondeterministic outcomes of actions. While formal methods claim to manage complexity, provide safety guarantees, and ensure robustness, empirical evaluation is necessary to validate these claims.This work presents an empirical study comparing the characteristics of various ad hoc behavior planning implementations (developed by participants with diverse levels of experience in computer science), with implementations using Tumato. We investigate the usability of the different approaches and evaluate i) their effectiveness, ii) the achieved safety (guarantees), iii) their robustness in handling uncertainties, and iv) their adaptability, extensibility, and scalability. To our knowledge, this is the first participant-based empirical study of a formal approach for (safe and robust) autonomous behavior.Our analysis confirms that while ad hoc methods offer some development flexibility, they lack the rigorous safety guarantees provided by formal methods. The study supports the hypothesis that formal methods, as implemented in Tumato, are effective tools for developing safe autonomous systems, particularly in managing complexity and ensuring robust decision making and planning.

Comparative Analysis of Value Iteration and Policy Iteration in Robotic Decision-Making

This study focuses on Markov Decision Processes (MDPs) as a framework for decision-making in robotics. MDPs are crucial for enabling autonomous systems to operate in dynamic environments. MDPs enable the development of optimal strategies that balance immediate and future rewards, making them essential for intelligent robotic behavior. The purpose of this study is to evaluate the performance differences of value iteration and strategy iteration algorithms in different robot tasks. The study highlights the strengths and weaknesses of each algorithm. This comparative analysis uniquely addresses a gap in existing research by evaluating both Value Iteration and Policy Iteration side by side, offering critical insights into their respective performances across diverse robotic tasks. Results indicate that Policy Iteration converges faster and adapts better in complex environments, making it ideal for real-time applications, while Value Iteration is more efficient in smaller state spaces.These findings provide critical insights for selecting algorithms tailored to specific robotic applications, highlighting their role in advancing intelligent robotic systems.

A Bio-Inspired Magnetic Soft Robotic Fish for Efficient Solar-Energy Driven Water Purification.

Solar-driven water evaporation is a promising solution for global water scarcity but is still facing challenges due to its substantial energy requirements. Here, a magnetic soft robotic bionic fish is developed by combining magnetic nanoparticles (Fe3O4), poly(N-isopropylacrylamide), and carboxymethyl chitosan. This bionic fish can release liquid water through hydrophilic/hydrophobic phase transition and dramatically reduce energy consumption. The introduced Fe3O4 nanoparticles endow the bionic fish with magnetic actuation capability, allowing for remote operation and recovery. Additionally, the magnetic actuation process accelerates the water absorption rate of the bionic fish as confirmed by the finite element simulations. The results demonstrate that bionic fish can effectively remove not only organic molecular dyes dissolved in water but also harmful microbes and insoluble microparticles from natural lakes. Moreover, the bionic fish maintains a good purification efficiency even after five recycling cycles. Furthermore, the bionic fish possesses other functions, such as salt purification and salt rejection. Finally, the mechanism of water purification is explained in conjunction with molecular dynamics calculations. This work provides a new approach for efficient solar-energy water purification by phase transition behavior in soft robotics.

Understanding the Interaction between Delivery Robots and Other Road and Sidewalk Users: A Study of User-generated Online Videos

The deployment of autonomous delivery robots in urban environments presents unique challenges in navigating complex traffic conditions and interacting with diverse road and sidewalk users. Effective communication between robots and road and sidewalk users is crucial to address these challenges. This study investigates real-world encounter scenarios where delivery robots and road and sidewalk users interact, seeking to understand the essential role of communication in ensuring seamless encounters. Following an online ethnography approach, we collected 117 user-generated videos from TikTok and their associated 2,067 comments. Our systematic analysis revealed several design opportunities to augment communication between delivery robots and road and sidewalk users, which include facilitating multi-party path negotiation, managing unexpected robot behaviour via transparency information, and expressing robot limitations to request human assistance. Moreover, the triangulation of video and comments analysis provides a set of design considerations to realise these opportunities. The findings contribute to understanding the operational context of delivery robots and offer insights for designing interactions with road and sidewalk users, facilitating their integration into urban spaces.

Development of a Robotic Device That Performs Head Bunting to Relieve User Tension

Inspired by an animal behavior called bunting, in which the animal rubs its head against other objects, including humans, we developed robotic prototypes capable of performing such bunting behaviors. Since physical contact plays an important role in therapeutic interactions between pets and their owners, we hypothesize that robot bunting can have a similar effect on its user. This paper reports on the development of such a robot with a flexible neck that can change its stiffness while the robot is rubbing its head against the user. An exploratory study was also conducted with 22 human participants, on whom the developed robot performed head bunting with three different (low/high/variable) stiffness conditions. The results show that the participants’ psychological tension, as measured by the Temporary Mood Scale (TMS), was significantly reduced ( \(p\lt 0.001\) ) after interacting with the robot. The difference between the three stiffness conditions was not significant in this study. Due to the lack of a control condition, we cannot confirm a clear effect of the stiffness change; however, some participants commented that the stiffness change made the robot's behavior lifelike and relaxing.

Classification of Co-Manipulation Modus with Human-Human Teams for Future Application to Human-Robot Systems

Despite the existence of robots that can physically lift heavy loads, robots that can collaborate with people to move heavy objects are not readily available. This article makes progress toward effective human-robot co-manipulation by studying 30 human-human dyads that collaboratively manipulated an object weighing \(27 \mathrm{kg}\) without being co-located (i.e., participants were at either end of the extended object). Participants maneuvered around different obstacles with the object while exhibiting one of four modi–the manner or objective with which a team moves an object together–at any given time. Using force and motion signals to classify modus or behavior was the primary objective of this work. Our results showed that two of the originally proposed modi were very similar, such that one could effectively be removed while still spanning the space of common behaviors during our co-manipulation tasks. The three modi used in classification were quickly , smoothly and avoiding obstacles . Using a deep convolutional neural network (CNN), we classified three modi with up to 89% accuracy from a validation set. The capability to detect or classify modus during co-manipulation has the potential to greatly improve human-robot performance by helping to define appropriate robot behavior or controller parameters depending on the objective or modus of the team.

Diffusion policy: Visuomotor policy learning via action diffusion

This paper introduces Diffusion Policy, a new way of generating robot behavior by representing a robot’s visuomotor policy as a conditional denoising diffusion process. We benchmark Diffusion Policy across 15 different tasks from 4 different robot manipulation benchmarks and find that it consistently outperforms existing state-of-the-art robot learning methods with an average improvement of 46.9%. Diffusion Policy learns the gradient of the action-distribution score function and iteratively optimizes with respect to this gradient field during inference via a series of stochastic Langevin dynamics steps. We find that the diffusion formulation yields powerful advantages when used for robot policies, including gracefully handling multimodal action distributions, being suitable for high-dimensional action spaces, and exhibiting impressive training stability. To fully unlock the potential of diffusion models for visuomotor policy learning on physical robots, this paper presents a set of key technical contributions including the incorporation of receding horizon control, visual conditioning, and the time-series diffusion transformer. We hope this work will help motivate a new generation of policy learning techniques that are able to leverage the powerful generative modeling capabilities of diffusion models. Code, data, and training details are available (diffusion-policy.cs.columbia.edu).

Optimizing the grasping behavior of a soft robot's gripper

AbstractIn this study, smart composite materials based on 4D printing technology were used to determine the grasping ability of four‐claw gripper of soft robot. First, fused deposition modeling (FDM) is used to prepare polylactic acid (PLA) strips on the surface of a rectangular bamboo substrate to create a single‐claw gripper for a smart composite soft robot, while using stimulus (heat) to deform and recover its original shape. Here, the authors first evaluate the deformation and recovery angle of the single‐claw gripper of the smart composite soft robot, as well as various processing parameters such as heating temperature and time. This study then used FDM to print PLA ribbons on cross‐shaped bamboo sheets to design the four‐claw gripper of soft robot and its gripper technology was actuation type. Then, the four‐claw gripper of soft robot was used to grab the table tennis ball to test its grasping ability. Finally, four processing parameters (width of claw, pitch of PLA ribbon, heating temperature, and heating time) were applied to determine the grasping ability of the four‐claw gripper of soft robot. The optimal processing parameters for the grasping ability of the four‐claw gripper of soft robot were width of claw of 20.2 mm, pitch of PLA ribbon of 1 mm, heating temperature of 200°C, and the heating time of 15 s. The authors also applied the operation window of two processing parameters to discuss the success of the grasping ability of the four‐claw gripper of soft robot. The innovation of this study is the use of PLA/bamboo composite materials as the four‐claw gripper of soft robot and these materials are cheap, easy to obtain, and can be recycled naturally. This study uses a simple thermal stimulus to enable the four‐claw gripper of soft robot to easily grasp irregular objects.Highlights The innovative composite material of polylactic acid and bamboo fabricated by 3D printing. Width of claw and heating temperature are the key factors on grasping ability. Operating window of grasping ability appears on four‐claw gripper of soft robot.

Social robot detection based on user behavioral representation

Social robots aim to facilitate our lives by mimicking our behaviors to deliver information and interact with us. Unfortunately, some social robots are bred for sinister intention, which often spread fake news, even threatening personal and national security. To effectively detect these robots, some existing works utilize basic user attributes for clustering, such as creation time and other information, but ignore the behavioral features of social robots, which are important a priori information for detection, and their absence leads to poor performance. To address this problem, we construct the Social Robot Behavior Detection model, which employs social robots' behavioral habits in disseminating content. Specifically, our method consists of three parts. First, to accurately characterize news, we analyze news features from multiple views, which facilitates representing users' habits of posting news. Second, a novel aggregation strategy is designed for aggregating user features to obtain accurate user representation. Finally, a classifier is used to detect whether user is social robots or not. We conduct analytical experiments on two public datasets, and the results show that our model has better detection performance, which demonstrates that our approach outperforms existing works on social robot detection. Our source code is available at https://github.com/NPURG/SRDB.