Motion Planning for Multi-Mobile-Manipulator Payload Transport Systems

Tallamraju, Rahul; Salunkhe, Durgesh Haribhau; Rajappa, Sujit; Ahmad, Aamir; Karlapalem, Kamalakar; Shah, Suril V.

doi:10.1109/coase.2019.8842840

“…The continuous collision avoidance reward is given as follows. 3 is that here we use a potential field-based collision avoidance method [20] as a part of the environment during the training to keep the robots from colliding with each other at all times. It is not embedded in the reward structure and hence, the robots are not explicitly penalized for it.…”

Section: Proposed Methodologymentioning

confidence: 99%

AirCapRL: Autonomous Aerial Human Motion Capture Using Deep Reinforcement Learning

Tallamraju

¹

,

Saini

²

,

Bonetto

³

et al. 2020

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

In this letter, we introduce a deep reinforcement learning (DRL) based multi-robot formation controller for the task of autonomous aerial human motion capture (MoCap). We focus on vision-based MoCap, where the objective is to estimate the trajectory of body pose and shape of a single moving person using multiple micro aerial vehicles. State-of-the-art solutions to this problem are based on classical control methods, which depend on hand-crafted system and observation models. Such models are difficult to derive and generalize across different systems. Moreover, the non-linearities and non-convexities of these models lead to sub-optimal controls. In our work, we formulate this problem as a sequential decision making task to achieve the vision-based motion capture objectives, and solve it using a deep neural network-based RL method. We leverage proximal policy optimization (PPO) to train a stochastic decentralized control policy for formation control. The neural network is trained in a parallelized setup in synthetic environments. We performed extensive simulation experiments to validate our approach. Finally, real-robot experiments demonstrate that our policies generalize to real world conditions.

show abstract

“…where d lthresh = 1.0m and d hthresh = 20m. v pot is obtained using the potential field functions as described in our previous work [20] (equation 3). Furthermore, the value of v pot is clamped to 1. d) Network 2.4 + Potential Field: Centering and Multiview HMR Reward (Trained with Moving Subject): In this variant, we use a sum of two rewards, namely, r center (4) and r MHMR (9).…”

Section: Proposed Methodologymentioning

confidence: 99%

“…The key difference in this case w.r.t. Network 2.3 is that here we use a potential field-based collision avoidance method [20] as a part of the environment during the training to keep the robots from colliding with each other at all times. It is not embedded in the reward structure and hence, the robots are not explicitly penalized for it.…”

Section: Proposed Methodologymentioning

confidence: 99%

AirCapRL: Autonomous Aerial Human Motion Capture using Deep Reinforcement Learning

Tallamraju¹,

Saini²,

Bonetto³

et al. 2020

Preprint

Self Cite

0

View full text Add to dashboard Cite

In this letter, we introduce a deep reinforcement learning (DRL) based multi-robot formation controller for the task of autonomous aerial human motion capture (MoCap). We focus on vision-based MoCap, where the objective is to estimate the trajectory of body pose and shape of a single moving person using multiple micro aerial vehicles. State-of-the-art solutions to this problem are based on classical control methods, which depend on hand-crafted system and observation models. Such models are difficult to derive and generalize across different systems. Moreover, the non-linearities and non-convexities of these models lead to sub-optimal controls. In our work, we formulate this problem as a sequential decision making task to achieve the vision-based motion capture objectives, and solve it using a deep neural network-based RL method. We leverage proximal policy optimization (PPO) to train a stochastic decentralized control policy for formation control. The neural network is trained in a parallelized setup in synthetic environments. We performed extensive simulation experiments to validate our approach. Finally, real-robot experiments demonstrate that our policies generalize to real world conditions.

show abstract

“…A 2D version, called outlined rectangle, has been considered to obtain a convenient shape of the multirobot transportation system in previous studies [13], [14]. Similarly, the 2D deformable box is adopted in a motion planning strategy for approximating the shape of a team of manipulators and the transported rigid payload [15]. Another study about probabilistic roadmap motion planning for deformable robots obtains the deformation of the robot from a deformed bounding box [16].…”

Section: Introductionmentioning

confidence: 99%

Multirobot Transport of Deformable Objects With Collision Avoidance

Herguedas

¹

,

López‐Nicolás

²

,

Sagüés

³

2023

IEEE Systems Journal

2

0

View full text Add to dashboard Cite

In the problem of autonomous transport of deformable objects, we propose a multirobot approach for steering a large object to a target configuration (object dimensions, orientation and position). Firstly, we present a deformation model based on the evolution over time of the dimensions and rotation of the object's bounding box. We consider the object is grasped by a set of mobile robots with double-integrator dynamics. Then, we propose a set of nominal controllers that allows reaching a desired configuration of the bounding box that models the deformable object. In order to prevent collisions of the object with static or dynamic obstacles, we formulate a control barrier function (CBF) that exploits our deformation model. Finally, we integrate the nominal controllers and the CBF into a quadraticprogramming controller, which includes overstretching avoidance and velocity constraints. We report simulation results to show the performance of this approach in different test scenarios.

show abstract

Motion Planning for Multi-Mobile-Manipulator Payload Transport Systems

Cited by 10 publications

References 21 publications

AirCapRL: Autonomous Aerial Human Motion Capture Using Deep Reinforcement Learning

AirCapRL: Autonomous Aerial Human Motion Capture Using Deep Reinforcement Learning

AirCapRL: Autonomous Aerial Human Motion Capture using Deep Reinforcement Learning

Multirobot Transport of Deformable Objects With Collision Avoidance

Contact Info

Product

Resources

About