Towards advanced robotic manipulation for nuclear decommissioning: A pilot study on tele-operation and autonomy

Marturi, Naresh; Rastegarpanah, Alireza; Takahashi, Chie; Adjigble, Maxime; Stolkin, Rustam; Zurek, Sebastian; Kopicki, Marek; Talha, Mohammed; Kuo, Jeffrey A.; Bekiroglu, Yasemin

doi:10.1109/raha.2016.7931866

Cited by 66 publications

(48 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, robots are needed for sorting and segregation of huge quantities of hazardous nuclear waste, which can contain numerous kinds of complex deformable materials (contaminated protective suits, overalls, gloves, tarpaulins), and where recognition of individual objects is extremely important for judging the level of hazard (e.g. when sorting and inventorying the unknown contents of half-century old waste containers) [1].…”

Section: Introductionmentioning

confidence: 99%

Single-shot clothing category recognition in free-configurations with application to autonomous clothes sorting

Sun

Aragon-Camarasa

Rogers

et al. 2017

2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Self Cite

View full text Add to dashboard Cite

Abstract-This paper proposes a single-shot approach for recognising clothing categories from 2.5D features. We propose two visual features, BSP (B-Spline Patch) and TSD (Topology Spatial Distances) for this task. The local BSP features are encoded by LLC (Locality-constrained Linear Coding) and fused with three different global features. Our visual feature is robust to deformable shapes and our approach is able to recognise the category of unknown clothing in unconstrained and random configurations. We integrated the category recognition pipeline with a stereo vision system, clothing instance detection, and dual-arm manipulators to achieve an autonomous sorting system. To verify the performance of our proposed method, we build a high-resolution RGBD clothing dataset of 50 clothing items of 5 categories sampled in random configurations (a total of 2,100 clothing samples). Experimental results show that our approach is able to reach 83.2% accuracy while classifying clothing items which were previously unseen during training. This advances beyond the previous state-of-the-art by 36.2%. Finally, we evaluate the proposed approach in an autonomous robot sorting system, in which the robot recognises a clothing item from an unconstrained pile, grasps it, and sorts it into a box according to its category. Our proposed sorting system achieves reasonable sorting success rates with single-shot perception.

show abstract

Section: Introductionmentioning

confidence: 99%

Single-shot clothing category recognition in free-configurations with application to autonomous clothes sorting

Sun

Aragon-Camarasa

Rogers

et al. 2017

2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Self Cite

View full text Add to dashboard Cite

show abstract

“…All the participants had a normal to corrected vision. Previously developed software [16] has been used to interface the robot motion control with a gaming joypad, which allows the operator to switch between and jog the robot in different frames (joint, base and tool) as well as to control the attached tool, i.e. a two finger parallel jaw gripper.…”

Section: Data Acquisition For Performance Analysismentioning

confidence: 99%

“…For the sake of analysing the performance of a semi-autonomous system as well to compare the human performance in case of stacking objects, a simple position-based visual servoing scheme has been developed as in Ref. [16] to automatically manoeuvre the robot to a desired grasping location and to stack objects. A trivial visual control law has been used in combination with a model-based 3D pose matching [19].…”

Section: Semi-autonomous Systemsmentioning

confidence: 99%

See 1 more Smart Citation

Towards Advanced Robotic Manipulations for Nuclear Decommissioning

Marturi¹,

Rastegarpanah²,

Rajasekaran³

et al. 2017

Robots Operating in Hazardous Environments

Self Cite

View full text Add to dashboard Cite

Despite enormous remote handling requirements, remarkably very few robots are being used by the nuclear industry. Most of the remote handling tasks are still performed manually, using conventional mechanical master-slave devices. The few robotic manipulators deployed are directly tele-operated in rudimentary ways, with almost no autonomy or even a pre-programmed motion. In addition, majority of these robots are under-sensored (i.e. with no proprioception), which prevents them to use for automatic tasks. In this context, primarily this chapter discusses the human operator performance in accomplishing heavy-duty remote handling tasks in hazardous environments such as nuclear decommissioning. Multiple factors are evaluated to analyse the human operators' performance and workload. Also, direct human tele-operation is compared against human-supervised semi-autonomous control exploiting computer vision. Secondarily, a vision-guided solution towards enabling advanced control and automating the undersensored robots is presented. Maintaining the coherence with real nuclear scenario, the experiments are conducted in the lab environment and results are discussed.

show abstract

“…Teleoperation is enabling robotic systems to become pervasive in settings where full autonomy is currently out of reach [18], [47], [15]. Compelling applications include minimally invasive surgery [44], [45], space exploration [46], remote vehicle operations [16] and disaster relief scenarios [34]. A human teleoperator can control a robot through tasks that have complex semantics and are currently difficult to explicitly program or to learn to solve efficiently without supervision.…”

Section: Introductionmentioning

confidence: 99%

Teleoperator Imitation with Continuous-Time Safety

Khadir

Varley

Sindhwani

2019

Robotics: Science and Systems XV

View full text Add to dashboard Cite

Learning to effectively imitate human teleoperators, with generalization to unseen and dynamic environments, is a promising path to greater autonomy enabling robots to steadily acquire complex skills from supervision. We propose a new motion learning technique rooted in contraction theory and sum-of-squares programming for estimating a control law in the form of a polynomial vector field from a given set of demonstrations. Notably, this vector field is provably optimal for the problem of minimizing imitation loss while providing continuous-time guarantees on the induced imitation behavior. Our method generalizes to new initial and goal poses of the robot and can adapt in real-time to dynamic obstacles during execution, with convergence to teleoperator behavior within a well-defined safety tube. We present an application of our framework for pick-and-place tasks in the presence of moving obstacles on a 7-DOF KUKA IIWA arm. The method compares favorably to other learning-from-demonstration approaches on benchmark handwriting imitation tasks.

show abstract

Towards advanced robotic manipulation for nuclear decommissioning: A pilot study on tele-operation and autonomy

Cited by 66 publications

References 19 publications

Single-shot clothing category recognition in free-configurations with application to autonomous clothes sorting

Single-shot clothing category recognition in free-configurations with application to autonomous clothes sorting

Towards Advanced Robotic Manipulations for Nuclear Decommissioning

Teleoperator Imitation with Continuous-Time Safety

Contact Info

Product

Resources

About