The main goal of this thesis work is to ensure safe interaction between a human and a collaborative robot relying on a combination of trajectory planning and artificial potential field-based algorithms to control the target reaching and avoiding collision with an external dynamic obstacle. In the thesis, a general introduction to Robotics and the fundamentals of a collaborative robot are given. Then, the theoretical background of trajectory planning and collision avoidance, instrumental for the development of the control algorithms, is provided. Then, the thesis includes a comprehensive section regarding the collaborative robot which was been used in this thesis work, i.e., the Franka-Emika Panda manipulator, also providing a detailed overview of the manipulator in the simulation environment. The proposed collision avoidance scheme has been assessed in simulation. In particular, the robot was required to reach a certain target while avoiding a dynamic obstacle which moves in the environment, first following a predefined path unknown to the controller and then randomly. Finally, a modified version of the developed algorithm has been tested experimentally. In particular, an Xbox 360 Kinect has been used to represent a human operator as a dynamic obstacle that the robot must avoid.
Sviluppo e attuazione di algoritmi per interazione uomo-robot sicura. L'obiettivo principale di questo lavoro di tesi è quello di combinare algoritmi di trajectory planning e di collision avoidance basati sulla tecnica degli artificial potential fields, al fine di garantire un'interazione sicura fra uomo e robot. Nella tesi, viene inizialmente fatta un'introduzione riguardo alla Robotica, presentando le caratterisctiche di un robot collaborativo. Dopodichè, viene introdotto il background teorico relativo al trajectory planning e alla collision avoidance, strumentale per lo sviluppo del'algoritmo di controllo. Inoltre, la tesi include una sezione che descrive il robot collaborativo utilizzato nelle simulazioni e negli esperimenti, i.e., il manipolatore Panda prodotto da Franka-Emika. La validità della metodologia di collision avoidance è stata verificata prima in simulazione. Nello specifico, il robot è stato controllato cosicchè raggiungesse un certo target, evitando un oggetto dinamico che si muove nell'ambiente, dapprima seguendo un path prestabilito ma non noto al controllore e poi in maniera casuale. Infine, una versione modificata dell'algoritmo di controllo è stata testata sperimentalmente, usando un Xbox 360 Kinect per rappresentare l'operatore umano come un ostacolo dinamico che il robot deve evitare.
Development and Implementation of algorithms for Human-robot safe interaction
THIRUMALAI, SRIHARAN
2021/2022
Abstract
The main goal of this thesis work is to ensure safe interaction between a human and a collaborative robot relying on a combination of trajectory planning and artificial potential field-based algorithms to control the target reaching and avoiding collision with an external dynamic obstacle. In the thesis, a general introduction to Robotics and the fundamentals of a collaborative robot are given. Then, the theoretical background of trajectory planning and collision avoidance, instrumental for the development of the control algorithms, is provided. Then, the thesis includes a comprehensive section regarding the collaborative robot which was been used in this thesis work, i.e., the Franka-Emika Panda manipulator, also providing a detailed overview of the manipulator in the simulation environment. The proposed collision avoidance scheme has been assessed in simulation. In particular, the robot was required to reach a certain target while avoiding a dynamic obstacle which moves in the environment, first following a predefined path unknown to the controller and then randomly. Finally, a modified version of the developed algorithm has been tested experimentally. In particular, an Xbox 360 Kinect has been used to represent a human operator as a dynamic obstacle that the robot must avoid.È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/15839