With a wide variety of products in the assembly line, it is necessary to supply the components through kitting operations, which consists in creating a set of parts to be sent to the assembly lines. This challenging logistical task is frequently performed manually by warehousemen. The work of taking the objects is often light but implies a great repetitiveness in arm movement. Workers perceive work as repetitive and have some physically stressful working situations that can lead to work-related muskuloschelatal disorders. This thesis propose a collaborative approach to kitting operations, i.e. some parts are collected by the warehouseman while others are picked by the robot. More specifically, an online scheduling algorithm has been developed that improves the cycle time of the overall system and reduce worker’s effort when assembling the kit. The algorithm generates tasks schedule in real time and sends information to the robot and human about the object to be taken. The ergonomics information are provided by a function obtained offline with the help of Kinect and the REBA method. Finally, the new solution was tested in realistic experimental human-robot kitting system.
Le linee di assemblaggio presentano grandi varietà di prodotti; ciò rende necessario fornire i componenti attraverso operazioni di kitting, che consistono nel creare un insieme di parti da inviare alla linea di assemblaggio. Questo compito, logisticamente complesso, viene solitamente assolto da operai. Pur non essendo un lavoro fisicamente dispendioso, prevede una gran ripetitività di movimenti delle braccia, e viene quindi percepito come stressante dai lavoratori. Questi fattori possono portare al sorgere di disturbi muscoloscheletrici. Questa tesi illustra un approccio al kitting collaborativo nel quale alcune parti che compongono il kit vengono prese dall'operaio mentre altre sono prese dal robot. In particolare, è stato sviluppato un algoritmo di scheduling online che diminuisce il tempo di preparazione del kit e al contempo riduce lo sforzo fisico percepito dall’operaio durante l’assemblaggio dello stesso. L’algoritmo genera un piano dei compiti in tempo reale e invia informazioni agli agenti circa quale oggetto debba essere preso. Le informazioni riguardo l’ergonomia di ogni azione di picking vengono fornite all'algoritmo da una funzione ottenuta offline combinando il metodo REBA e il Kinect. Infine, la nuova soluzione è stata testata in un sistema di kitting sperimentale.
CoKitting : an online scheduling algorithm for human-robot kitting collaboration
POGGIALI, MATTEO
2018/2019
Abstract
With a wide variety of products in the assembly line, it is necessary to supply the components through kitting operations, which consists in creating a set of parts to be sent to the assembly lines. This challenging logistical task is frequently performed manually by warehousemen. The work of taking the objects is often light but implies a great repetitiveness in arm movement. Workers perceive work as repetitive and have some physically stressful working situations that can lead to work-related muskuloschelatal disorders. This thesis propose a collaborative approach to kitting operations, i.e. some parts are collected by the warehouseman while others are picked by the robot. More specifically, an online scheduling algorithm has been developed that improves the cycle time of the overall system and reduce worker’s effort when assembling the kit. The algorithm generates tasks schedule in real time and sends information to the robot and human about the object to be taken. The ergonomics information are provided by a function obtained offline with the help of Kinect and the REBA method. Finally, the new solution was tested in realistic experimental human-robot kitting system.File | Dimensione | Formato | |
---|---|---|---|
2019_07_Poggiali.pdf
Open Access dal 12/07/2020
Descrizione: Testo della tesi
Dimensione
60.27 MB
Formato
Adobe PDF
|
60.27 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/148594