Towards Adaptive Interaction Strategies for Human Companion Robot via Deep Reinforcement Learning

Cong-Thanh Vu; Yen-Chen Liu

cs.ROAug 20, 2026

Towards Adaptive Interaction Strategies for Human Companion Robot via Deep Reinforcement Learning

Authors: Cong-Thanh Vu, Yen-Chen Liu

Organizations: Department of Mechanical Engineering, National Cheng Kung University, Tainan 70101, Taiwan

Abstract

In the field of Human-Robot Interaction (HRI), achieving flexibility in human-accompanying within real-world environments holds great potential for various applications but also poses significant challenges. Traditional methods typically restrict robots to fixed positions relative to humans, such as tracking from behind, in front, or side-by-side, which limits robot adaptability in dynamic workspaces. This study introduces a novel human-companioning strategy that uses Reinforcement Learning (DRL) to enable mobile robots to dynamically adjust their tracking positions according to varying conditions. An interaction space is defined to capture the relationship between the human and the robot while considering the environment, which serves as the basis for state spaces in DRL to assist the robot in adapting to environmental changes. A human-robot companion controller is developed by integrating Model Predictive Path Integral (MPPI) control with Control Barrier Functions (CBF), ensuring that the robot accurately follows the target's movement in both position and orientation while avoiding obstacles and enhancing social acceptance and safety. The proposed approach is evaluated in real-world scenarios, both indoors and outdoors, and compared with other studies. The results show that the proposed method improves the success rate and tracking accuracy by at least 24% and 47%, respectively, while enhancing human comfort. Experiments demonstrate the robot's ability to flexibly accompany a person walking at speeds of up to 1.7 m/s, dynamically adjusting its strategy without being confined to a fixed position. Additionally, the robot respects the human's intimate space to ensure safety, comfort, and effective obstacle avoidance.

Figures & tables

[1.35pt] Method	Predefined Tracking Point	Adaptive Companionship	Follow Behind	Follow Side-by-Side	Follow In-Front	Obstacle Avoidance	Social Space	Outdoor
[1.0pt] Proposed	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Corridor-Intersection Recognition [ 46 ]	Yes	No	Yes	Only in corridors	No	No	No	-
PID [ 2 ] , Gait Recog. [ 3 ] , Deep-MPC [ 7 ] , HFC [ 42 ]	Yes	No	Yes	No	No	No	No	-
Hedge Algebra Fuzzy [ 40 ] , DWA [ 5 ] , RPF-Search [ 47 ] , Deep-RL [ 29 ]	Yes	No	Yes	No	No	Yes	No	-
Adaptive SFM [ 34 ] , MPC [ 31 ] , LQR [ 39 ] , Extended SFM [ 33 ]	Yes	No	No	Yes	No	Yes	No	-
Dual Closed-Loop [ 30 ]	Yes	No	No	Yes	No	Yes	Yes	-

TABLE I : Comparison with Related Works

Fig. 1 : Overview of the human-interactive robot.

Fig. 2 : Schematic model and generalized coordinates.

Fig. 3 : The overall structure of the system architecture. The robot observes the proposed cost functions, total cost, collision-free space, robot velocity, human position and velocity, obstacles, and previous actions. Policy determines an action index within the collision-free space and then provides a target position for the human-companion controller.

Fig. 4 : Illustration of the training environment.

Fig. 5 : Policy performance under varying hyperparameter configurations: (a) Proposed architecture and (b) RL baseline.

Fig. 6 : Simulation environment used for comparison.

Metrics	SR (%)	UT (s)	$e_{x}$ (m)	$e_{y}$ (m)	$e_{\theta}$ (rad)
PID [ 2 ]	0	-	-	-	-
DWA [ 5 ]	34	$33.0\pm 8.5$	$0.97\pm 0.25$	$0.88\pm 0.22$	$0.71\pm 0.18$
MPC [ 31 ]	46	$21.0\pm 5.4$	$0.41\pm 0.13$	$0.43\pm 0.14$	$0.23\pm 0.06$
DRL [ 29 ]	62	$27.0\pm 6.3$	$0.17\pm 0.07$	$0.21\pm 0.11$	$0.20\pm 0.08$
Proposed	86	$\mathbf{2.5\pm 0.8}$	$\mathbf{0.09\pm 0.03}$	$\mathbf{0.05\pm 0.02}$	$\mathbf{0.08\pm 0.03}$

TABLE II : Quantitative Results of Comparison.

Fig. 7 : Distribution of robot positions relative to the human across four experimental scenarios.

Fig. 8 : Tracking performance and user scores under different interaction-space configurations.

Fig. 9 : Tracking performance of the proposed controller in indoor scenarios: relative positions of the following point to the human ( $x_{d}^{h}$ , $y_{d}^{h}$ ) and the robot ( $x_{d}^{r}$ , $y_{d}^{r}$ ), tracking errors ( $e_{x}$ , $e_{y}$ , $e_{\theta}$ ), and the distance between the human and the robot ( $d_{hr}$ ).

Fig. 10 : Trajectories of the human and robot in four indoor scenarios.

Fig. 11 : Human and robot trajectories in indoor occlusion scenarios (doorway and multi-pedestrian).

Method	Success Rate (%)	Frequency (Hz)	Lighting Robustness	FoV
Camera	55	$\approx 30$	Low	$110^{\circ}$
LiDAR	39	$\approx 9$	High	$360^{\circ}$
AmfiTrack	92	$\approx\mathbf{120}$	High	Omni

TABLE III : Quantitative Performance Comparison of Tracking Hardware Systems.

Fig. 12 : Tracking accuracy at different walking speeds.

Fig. 13 : Tracking performance under various obstacle conditions.

Fig. 14 : Outdoor experimental environments.

Fig. 15 : Tracking performance on asphalt and brick roads.

Pedestrians	SR (%)	UT (s)	APT (ms)
5	81	1.35	9.86
10	65	2.79	10.01
15	33	7.13	10.93
20	26	9.22	11.85

TABLE IV : Quantitative Performance across Crowd Densities.

Human-Robot Interaction

Human-robot interaction studies how people and robots communicate, coordinate, and safely work or live together.

See all 452 papers

Jul 1, 2026

Adaptive Companionship for Group-Following Robots: Handling Dynamically Changing Group Formations

Accompanying a group of humans is an essential aspect of developing human-like social cognition in robots. However, human groups typically do not follow fixed formations, which poses significant challenges for robots in maintaining natural companionship behaviors. In this paper, we propose an adaptive group-accompaniment method for social robots based on Vision-Language Models (VLMs), leveraging their semantic reasoning capabilities to infer companion positions, maintain social distances, and understand group dynamics. The members of the group are first detected, and a perceptual module generates visual representations of the interaction group space as input to the VLM, which is then combined with a Model Predictive Path Integral (MPPI) controller to ensure stability and safety. Experimental evaluations across five scenarios show that the proposed method enables robots to accompany the group effectively, demonstrating a 15% improvement in success rate and a 25% reduction in collision rate compared to baseline approaches. Additionally, a user study indicates that the generated companionship behaviors are perceived as natural and socially appropriate.

Cong-Thanh Vu, Yen-Chen Liu

Department of Mechanical Engineering, National Cheng Kung University, Tainan 70101, Taiwan.

Social Robot Navigation Human-Robot Interaction

Ask PDF

Jul 22, 2026

Gimbal-Based Human Tracking for Companion Robots Using Continual Learning

Reliable and continuous human tracking is essential for natural human-robot interaction, particularly for companion robots. However, many existing approaches rely on wearable tags or fixed cameras with limited fields of view, which reduces system flexibility and often causes tracking failures when the target moves outside the sensing range. In this paper, we present a human tracking approach based on a gimbal-mounted camera integrated into a mobile robot. By actively controlling the gimbal mechanism, the camera can dynamically adjust its viewing direction to maintain the target within the field of view, even under substantial relative motion between the robot and the human. Furthermore, a continual learning strategy is applied to the person re-identification (ReID) task to adapt to changes in appearance and environmental conditions during long-term tracking. Experimental results demonstrate that the proposed system significantly improves the stability and continuity of human tracking, enables real-time re-identification, and provides responsive feedback for reliable tracking of human motion from walking to running. User studies further indicate that the proposed approach enhances user comfort by eliminating the need for wearable tags.

Cong-Thanh Vu, Ching-Chieh Liu, Yen-Chen Liu

Department of Mechanical Engineering, National Cheng Kung University (NCKU), Tainan 70101, Taiwan

Visual Object Tracking Human-Robot Interaction Person Re-Identification

Ask PDF

Apr 24, 2026

Adaptive vs. Static Robot-to-Human Handover: A Study on Orientation and Approach Direction

Robot-to-human handovers often rely on static, open-loop strategies (or, at best, approaches that adapt only the position), which generally do not consider how the object will be grasped by the human, thus requiring the user to adapt. This work presents a novel adaptive framework that dynamically adjusts the object's delivery pose in real time based on the user's hand pose and the intended downstream task. By integrating AI-based hand pose estimation with smooth, kinematically constrained trajectories, the system ensures a safe approach and an optimal handover orientation. A comprehensive user study compares the proposed adaptive approach against a static baseline across multiple tasks, evaluating both subjective metrics (NASA-TLX, Human-Robot Trust Scale) and objective physiological data (blink rate measured via wearable eye-trackers). The results demonstrate that dynamic alignment significantly reduces users' cognitive workload and physiological stress, while improving their confidence in the robot's reliability. These findings highlight the potential of task- and pose-aware systems for enabling fluid and ergonomic human-robot collaboration.

Federico Biagi, Dario Onfiani, Simone Silenzi +2

Department of Engineering "Enzo Ferrari", University of Modena and Reggio Emilia, Italy. · Department of Surgery, Medicine, Dentistry and Morphological Sciences, University of Modena and Reggio Emilia, Italy.

Human-Robot Collaboration Human-Robot Interaction Robot Motion Planning

Ask PDF

Abstract

Human-Robot Interaction

Explore similar work

Adaptive Companionship for Group-Following Robots: Handling Dynamically Changing Group Formations

Gimbal-Based Human Tracking for Companion Robots Using Continual Learning

Adaptive vs. Static Robot-to-Human Handover: A Study on Orientation and Approach Direction