Decision-making algorithms for safe robotic disassembling of randomly piled objects. (17th December 2017)

• Record Type: Journal Article
• Authors: Temtsin, Sharon; Degani, Amir
• Is Part Of: Advanced robotics. Volume 31:Number 23/24(2018)
• Page Start: 1281
• Page End: 1295
• Publication Date: 2017-12-17
• Subjects: Pile disassembling -- autonomous robotics -- robotic mobile manipulation -- manipulation under uncertainties; Robotics -- Periodicals; Robotics -- Japan -- Periodicals; Robotics; Japan; Periodicals; 629.89205
• DOI: 10.1080/01691864.2017.1352537
• Languages: English
• ISSNs: 0169-1864

Abstract

Abstract: The task of disassembling a pile of objects in Urban Search and Rescue (USAR) sites often requires taking into consideration pile disturbances in order to keep the safety of the survivors, the USAR teams, or the safety of the site's environment. This work compares four decision-making algorithms (DMAs) for the autonomous pile disassembling process. The DMA must elect a single block for removal. The Volume Height Algorithm (VHA) removes the object with the highest potential energy; the Filtered Volume Height Algorithm (FVHA) removes the object with the highest potential energy that does not support any other objects from below; the Kinematic Based Algorithm (KBA) decides to remove the least encapsulated object; and lastly, the Filtered Volume Height Kinematic Based Algorithm (FVHKBA) is a combination of FVHA and KBA. All algorithms were simulated and analysed under full knowledge conditions and the more realistic, partial knowledge conditions. In partial knowledge mode, the algorithm processes the knowledge of only the objects which can be seen from outside viewing points. Additionally, the effect of the pile homogeneity was examined by disassembling piles of identical blocks and of different sized blocks. The FVHA shows the most impressive performances regardless of knowledge conditions and pile homogeneity. Abstract :