-
Learning Joint Space Reference Manifold for Reliable Physical Assistance
Authors:
Amirreza Razmjoo,
Tilen Brecelj,
Kristina Savevska,
Aleš Ude,
Tadej Petrič,
Sylvain Calinon
Abstract:
This paper presents a study on the use of the Talos humanoid robot for performing assistive sit-to-stand or stand-to-sit tasks. In such tasks, the human exerts a large amount of force (100--200 N) within a very short time (2--8 s), posing significant challenges in terms of human unpredictability and robot stability control. To address these challenges, we propose an approach for finding a spatial…
▽ More
This paper presents a study on the use of the Talos humanoid robot for performing assistive sit-to-stand or stand-to-sit tasks. In such tasks, the human exerts a large amount of force (100--200 N) within a very short time (2--8 s), posing significant challenges in terms of human unpredictability and robot stability control. To address these challenges, we propose an approach for finding a spatial reference for the robot, which allows the robot to move according to the force exerted by the human and control its stability during the task. Specifically, we focus on the problem of finding a 1D manifold for the robot, while assuming a simple controller to guide its movement on this manifold. To achieve this, we use a functional representation to parameterize the manifold and solve an optimization problem that takes into account the robot's stability and the unpredictability of human behavior. We demonstrate the effectiveness of our approach through simulations and experiments with the Talos robot, showing robustness and adaptability.
△ Less
Submitted 15 August, 2023;
originally announced January 2024.
-
The JSI-KneExo: Active, Quasi-Passive, Pneumatic, Portable Knee Exo with Bidirectional Energy Flow for Air Recovery in Sit-Stand Tasks
Authors:
Luka Mišković,
Tilen Brecelj,
Miha Dežman,
Tadej Petrič
Abstract:
While existing literature encompasses exoskeleton-assisted sit-stand tasks, the integration of energy recovery mechanisms remains unexplored. To push these boundaries further, this study introduces a portable pneumatic knee exoskeleton that operates in both quasi-passive and active modes, where active mode is utilized for aiding in standing up (power generation), thus the energy flows from the exo…
▽ More
While existing literature encompasses exoskeleton-assisted sit-stand tasks, the integration of energy recovery mechanisms remains unexplored. To push these boundaries further, this study introduces a portable pneumatic knee exoskeleton that operates in both quasi-passive and active modes, where active mode is utilized for aiding in standing up (power generation), thus the energy flows from the exoskeleton to the user, and quasi-passive mode for aiding in sitting down (power absorption), where the device absorbs and can store energy in the form of compressed air, leading to energy savings in active mode. The absorbed energy can be stored and later reused without compromising exoskeleton transparency in the meantime. In active mode, a small air pump inflates the pneumatic artificial muscle (PAM), which stores the compressed air, that can then be released into a pneumatic cylinder to generate torque. All electronic and pneumatic components are integrated into the system, and the exoskeleton weighs 3.9 kg with a maximum torque of 20 Nm at the knee joint. The paper describes the mechatronic design, mathematical model and includes a pilot study with an able-bodied subject performing sit-to-stand tasks. The results show that the exoskeleton can recover energy while assisting the subject and reducing muscle activity. Furthermore, results underscore air regeneration's impact on energy-saving in portable pneumatic exoskeletons.
△ Less
Submitted 6 September, 2023; v1 submitted 28 June, 2023;
originally announced June 2023.
-
Increased Complexity of a Human-Robot Collaborative Task May Increase the Need for a Socially Competent Robot
Authors:
Rebeka Kropivšek Leskovar,
Tadej Petrič
Abstract:
An important factor in develo** control models for human-robot collaboration is how acceptable they are to their human partners. One such method for creating acceptable control models is to attempt to mimic human-like behaviour in robots so that their actions appear more intuitive to humans. To investigate how task complexity affects human perception and acceptance of their robot partner, we pro…
▽ More
An important factor in develo** control models for human-robot collaboration is how acceptable they are to their human partners. One such method for creating acceptable control models is to attempt to mimic human-like behaviour in robots so that their actions appear more intuitive to humans. To investigate how task complexity affects human perception and acceptance of their robot partner, we propose a novel human-based robot control model for obstacle avoidance that can account for the leader-follower dynamics that normally occur in human collaboration. The performance and acceptance of the proposed control method were evaluated using an obstacle avoidance scenario in which we compared task performance between individual tasks and collaborative tasks with different leader-follower dynamics roles for the robotic partner. The evaluation results showed that the robot control method is able to replicate human behaviour to improve the overall task performance of the subject in collaboration. However, regarding the acceptance of the robotic partner, the participants' opinions were mixed. Compared to the results of a study with a similar control method developed for a less complex task, the new results show a lower acceptance of the proposed control model, even though the control method was adapted to the more complex task from a dynamic standpoint. This suggests that the complexity of the collaborative task at hand increases the need not only for a more complex control model but also a more socially competent control model.
△ Less
Submitted 11 July, 2022;
originally announced July 2022.
-
Leader-Follower Dynamics in Complex Obstacle Avoidance Task
Authors:
Rebeka Kropivšek Leskovar,
Jernej Čamernik,
Tadej Petrič
Abstract:
A question that many researchers in social robotics are addressing is how to create more human-like behaviour in robots to make the collaboration between a human and a robot more intuitive to the human partner. In order to develop a human-like collaborative robotic system, however, human collaboration must first be better understood. Human collaboration is something we are all familiar with, howev…
▽ More
A question that many researchers in social robotics are addressing is how to create more human-like behaviour in robots to make the collaboration between a human and a robot more intuitive to the human partner. In order to develop a human-like collaborative robotic system, however, human collaboration must first be better understood. Human collaboration is something we are all familiar with, however not that much is known about it from a kinematic standpoint. One dynamic that hasn't been researched thoroughly, yet naturally occurs in human collaboration, is for instance leader-follower dynamics. In our previous study, we tackled the question of leader-follower role allocation in human dyads during a collaborative reaching task, where the results implied that the subjects who performed higher in the individual experiment would naturally assume the role of a leader when in physical collaboration. In this study, we build upon the leader-follower role allocation study in human dyads by observing how the leader-follower dynamics change when the collaborative task becomes more complex. Here, the study was performed on a reaching task, where one subject in a dyad was faced with an additional task of obstacle avoidance when performing a 2D reaching task, while their partner was not aware of the obstacle. We have found that subjects change their roles throughout the task in order to complete it successfully, however looking at the overall task leader the higher-performing individual will always dominate over the lower-performing one, regardless of whether they are aware of the additional task of obstacle avoidance or not.
△ Less
Submitted 11 July, 2022;
originally announced July 2022.
-
Towards open-ended evolutionary simulator for develo** novel tumour drug delivery systems
Authors:
Igor Balaz,
Tara Petric,
Namid Stillman
Abstract:
Tumours behave as moving targets that can evade chemotherapeutic treatments by rapidly acquiring resistance via various mechanisms. In Balaz et al. (2021, Biosystems; 199:104290) we initiated the development of the agent-based open-ended evolutionary simulator of novel drug delivery systems (DDS). It is an agent-based simulator where evolvable agents can change their perception of the environment…
▽ More
Tumours behave as moving targets that can evade chemotherapeutic treatments by rapidly acquiring resistance via various mechanisms. In Balaz et al. (2021, Biosystems; 199:104290) we initiated the development of the agent-based open-ended evolutionary simulator of novel drug delivery systems (DDS). It is an agent-based simulator where evolvable agents can change their perception of the environment and thus adapt to tumour mutations. Here we mapped the parameters of evolvable agent properties to the realistic biochemical boundaries and test their efficacy by simulating their behaviour at the cell scale using the stochastic simulator, STEPS. We show that the shape of the parameter space evolved in our simulator is comparable to those obtained by the rational design.
△ Less
Submitted 25 May, 2021;
originally announced May 2021.
