Prepreg is a fibre reinforced polymer composite material often used by the aeronautical industry. The material supplier has pre-impregnated the fibre often with an epoxy resin and cured it to a semi-viscous B-stage where the material is tacky (adhesive). Manual layup of prepreg components is still common because there are only a few automated layup techniques which has limited geometric capability and high investment cost. Swerea SICOMP is a research institute which have in collaboration with partners developed a manufacturing demonstrator for automated layup of carbon fibre prepreg components based on robotic handling. The manufacturing demonstrator is able to; feed out prepreg from a spool onto a cutting table, cut the prepreg into plies, pick the plies from the cutting table and place them on a layup table, consolidate the prepreg plies on the layup table and remove the backing paper that covers one side of the prepreg. Three robustness problems has been identified and these forms the bulk of this master’s thesis. The first problem is that the robot is unable to place the plies on the cutting table with sufficient accuracy due to fluctuating position of the prepreg on the cutting table. The second problem is that the end effector used for the pick and place operation sometimes fails to pick the adhesive plies from the cutting table. The problem originates in limited holding force by the end effectors suction cups and limited capability to perform a peeling motion when picking due to the design of the end effector. The third problem is that process variables like temperature is believed to effect the robustness of the demonstrator but has not been further studied. The first problem was solved by implementing a probing routine which measures the position of the prepreg on the cutting table with an array sensor mounted to the robot. The measurement values are used to reprogram subsequent operations. The second problem seems to be solved by designing and implementing a new pick and place end effector. The new end effector has higher holding force and better capability to perform a peeling motion while picking and placing the tacky prepreg plies. Tests with the new end effector has been promising, the pick and placeoperation performs robustly, but all ply geometries was not tested and some minor adjustment is needed to further improve the placing accuracy. The third problem has not been solved but studied. A literature study reviled that temperature, relative humidity and prepreg out of freezer age is likely to effect the prepreg adhesive properties which in turn is known to effect the robustness. A test procedure has been devised to test how these variables effect the robustnessof the demonstrator. The tests should be conducted when all process steps in the demonstrator is up and running.

The aim of this thesis is to automate the depowdering step in binder-jet additive manufacturing using KUKA kr6 robot. The major tasks involved in this thesis work are commissioning of the KUKA kr6 robot, plan the actions required for automation process of binder jet considering the scaling factor of green bodies. For this purpose KUKA robot with a standard KUKA compact controller (KRC4) and KUKA system software 8.3 (KSS) has been used. In Peter Corke Matlab toolbox (Matlab toolbox), KUKA kr6 robot model has been simulated to understand forward kinematics problem which shows the study the motion of end effector of robot in space for picking process. These transformations between the Joint coordinate systems and Cartesian coordinate systems give the forward and inverse kinematics. Firstly, a KUKA kr6 robot has been programmed in a KUKA Robot Language (KRL) using an algebraic approach with geometric operator to automate the picking process of green bodies. These are fragile bodies thereby many number of tests have been conducted to improve the program. Also additional effort has been placed to test the customized gripper used for the layered pick and place of the components, customized vacuum cleaner for the vacuum cleaning in between the layers and to consider the scaling factor during the sintering step of the binder jet. Finally, KUKA kr6 robot model has been simulated in Matlab toolbox. The picking point in the space has been simulated to study the forward kinematics and to understand how the robot reaches a position and orientation in space. Cartesian trajectory has been simulated. Also, the Jacobian matrix, its rank and determinant are studied to understand the singularities in KUKA kr6 robot, basing on which the thesis work can be continued and enhanced further.

The objective of this master thesis is to design a robotic work cell for automated insertion of thermostatic heads. The proposed design was based on the comprehensive analysis of the current workplace. By taking the demands of the company and customer into consideration, several possibilities of the robotic work cell design were proposed, from which the most suitable one was chosen and carefully elaborated. During the process of designing, several safety standards had to be taken into consideration in order to mitigate risks. For that purpose, a risk analysis was conducted as well. Using the Siemens Process Simulate software, the proposed robotic work cell concept was verified, including cycle time analysis. The last part of the thesis is an assessment of the initial expenditures of the robotic work cell equipment, as well as the return of the investment calculation. Drawing of the robotic work cell layout is attached to the thesis.

A fundamental aspect to consider in factories is the transportation of the items at differentsteps in the production process. Conveyor belts do a great to bring items from point A topoint B but to load the item onto a working station it can demands a more precise and,in some cases, delicate approach. Nowadays this part is mostly handled by robotic arms.The issue encountered is that a robot arm extremity, its gripper, cannot directly instinctivelyknow how to grip an object. It is usually up to a technician to configure how andwhere the gripper goes to grip an item.The goal of this thesis is to analyse a problem given by a company which is to find a wayto automate the grasp pose synthesis of a new object with the adapted gripper.This automatized process can be separated into two sub-problems.First, how to choose the adapted gripper for a new object.Second, how to find a grasp pose on the object, with the previously chosen gripper.In the problem given by the company, the computer-aided design (CAD) 3D model of theconcerned object is given. Also, the grasp shall always be done vertically, i.e., the grippercomes vertically to the object and the gripper does not rotate on the x and y axis. Thegripper for a new object is selected between two kinds of grippers: two-finger paralleljawgripper and three-finger parallel-jaw gripper. No dataset of objects is provided.Object grasping is a well researched subject, especially for 2 finger grippers. However,few research is done for the 3 finger grippers grasp pose synthesis, or for gripper comparison,which are key part of the studied problem.To answer the sub-problems mentioned above, machine learning will be used for the gripperselection and a grasp synthesis method will be used for the grasp pose finding. However,due to the lack of gripper comparison in the related work, a new approach needsto be created, which will be inspired by the findings in the literature about grasp posesynthesis in general.This approach will consist of two parts.First, for each gripper and each object combination are generated some grasp poses, eachassociated with a corresponding score. The scores are used to have an idea of the bestgripper for an object, the best score for each gripper indicating how good a grasp couldbe on the object with said gripper.Secondly, the objects with their associated best score for each gripper will be used astraining data for a machine learning algorithm that will assist in the choice of the gripper.This approach leads to two research questions:“How to generate grasps of satisfying quality for an object with a certain gripper?”“Is it possible to determine the best gripper for a new object via machine learning ?”The first question is answered by using mathematical operations on the point cloud representationof the objects, and a cost function (that will be used to attribute a score), whileithe second question is answered using machine learning classification and regression togain insight on how machine learning can learn to associate object proprieties to gripperefficiency.The found results show that the grasp generation with the chosen cost function givesgrasp poses that are similar to the grasp poses a human operator would choose, but themachine learning models seem unable to assess grasp quality, either with regression orclassification.

[EN] The primary handling of food with robots calls for the development of new manipulation devices, especially when products are easily damaged and have a wide range of shapes and textures. These difficulties are even greater in the agricultural industry because the quality of the products is also checked during the manual handling process. This PhD dissertation provides solutions to these issues and helps to further introduce robotics into the handling of food. Several methods for handling food are included and analyzed, and specific solutions are proposed and then validated with prototypes. The research focuses on devices capable of adapting themselves to the shapes of the products without increasing the complexity of the mechanism. After analyzing several different solutions, the method chosen involves the use of under-actuated mechanisms, compliant mechanisms and fingers with pads filled with granular fluids. These fluids can behave as quasi-liquids or quasi-solids due to the jamming transition, which provides a soft initial grasp and can support high stresses during fast movements performed by the robot. The additive manufacturing process provides an opportunity to develop robot grippers that are lighter, simpler, more flexible and cheaper. By using this process elastic mechanisms are manufactured in a single part, which are equivalent to mechanisms with several rigid parts connected by joints. Laser sintering is employed to produce pneumatic actuators, with different types of motions, based on the elastic properties of the materials used in this manufacturing process. As a result, the systems can be simplified to achieve grippers, with several fingers, that are produced as a single part. In order to estimate the freshness and quality of agricultural products while they are being grasped, accelerometers are added to the fingers of several grippers. Accelerometers are economical and act as intrinsic tactile sensors. They can be easily embedded, thereby reducing the risk of getting damaged due to contact with the product, and allow each of the grasping phases to be identified. To achieve good performance of the accelerometers, a specific process is defined for the robot gripper, which touches the products a few times. In addition, several gripper prototypes are manufactured with diverse under-actuated mechanisms, jamming systems, and a new program that processes the signals from the accelerometers using different procedures in order to obtain parameters that can be used to estimate the quality of products. These parameters are correlated with data from destructive tests that are commonly used as a reference. The best performance of the accelerometers is achieved when the finger employs a granular fluid, a correlation coefficient of 0.937 being accomplished for the ripeness of mangoes and 0.872 for the firmness of eggplants. / [ES] La manipulación primaria de alimentos con robots precisa del desarrollo de nuevos sistemas de manipulación especialmente cuando los productos son sensibles al daño y presentan una amplia variabilidad de formas y texturas. En el sector agroalimentario las dificultades son aún mayores ya que la manipulación manual sirve además para inspeccionar los productos durante el proceso. Está tesis aporta soluciones a estos problemas facilitando la incorporación de la robótica. En la tesis se recopilan y analizan diversas soluciones para poder manipular alimentos proponiendo soluciones concretas que luego son validadas con prototipos. La investigación se centra en aquellos sistemas que son capaces de auto adaptarse a las formas de los productos sin incrementar la complejidad del mecanismo. Tras analizar diversas técnicas se propone el uso de mecanismos infra-actuados, mecanismos flexibles y dedos con fluidos granulares que, al estar encerrados dentro de una membrana, se comportan como cuasi-líquidos o cuasi-sólidos gracias a la transición jamming, permitiendo un agarre inicial suave y la posibilidad de transmitir esfuerzos elevados durante los movimientos del robot. En la búsqueda de garras más ligeras, sencillas, flexibles y económicas se aprovecha la oportunidad que brinda la tecnología de fabricación aditiva de material. Gracias a este proceso se fabrican mecanismos flexibles realizados en una única pieza y que equivalen a mecanismos de garras realizados con varias piezas rígidas unidos por articulaciones. Mediante el sinterizado por láser, se fabrican actuadores neumáticos, con diversos tipos de movimiento, basados en la flexibilidad del material empleado en su fabricación. En conjunto se simplifican los sistemas llegando a realizar garras flexibles de varios dedos fabricadas en una única pieza. Para evaluar la calidad y frescura de los productos agroalimentarios durante el agarre se emplean acelerómetros localizados en los dedos de varias garras. Los acelerómetros son económicos y se comportan como sensores táctiles intrínsecos, están fuera del contacto directo con el producto evitando desgastes por contacto y permiten identificar las distintas fases de agarre. Para lograr esto se desarrolla un proceso específico del robot con la garra, que palpa varias veces el producto. Se fabrican diversos tipos de garra con distintas tecnologías de mecanismos infra-actuados y sistemas jamming y se programa un algoritmo original de procesado de señal que con diversas técnicas es capaz de extraer parámetros de los acelerómetros que sirven para evaluar la calidad de los productos. Estos parámetros son correlacionados con los datos de ensayos destructivos que son habitualmente empleados como referencia. Las mejores capacidades se consiguen empleando garras con jamming lográndose coeficientes de correlación de 0.937 en índices de madurez con mangos y 0.872 en firmeza de berenjenas. / [CA] La manipulació primària d'aliments amb robots precisa del desenvolupament de nous sistemes de manipulació especialment quan els productes són sensibles al dany i presenten una àmplia variabilitat de formes i textures. En el sector agroalimentari les dificultats són encara més grans ja que la manipulació manual serveix a més per a inspeccionar els productes durant el procés. Aquesta tesi aporta solucions a aquests problemes facilitant la incorporació de la robòtica. En la tesi es recopilen i analitzen diverses solucions per a poder manipular aliments proposant solucions concretes que després són validades amb prototips. La investigació es centra en aquells sistemes que són capaços d'auto adaptar-se a la forma dels productes sense incrementar la complexitat del mecanisme. Després d'analitzar diverses tècniques es proposa l'ús de mecanismes infra-actuats, mecanismes flexibles i dits amb fluids granulars que, tancats dins d'una membrana, es comporten com quasi-líquids o quasi-sòlids gràcies a la transició jamming, permetent una prensió inicial suau i la possibilitat de transmetre esforços elevats durant els moviments del robot. En la recerca d'urpes més lleugeres, senzilles, flexibles i econòmiques s'aprofita l'oportunitat que brinda la tecnologia de fabricació additiva de material. Gràcies a aquest procés es fabriquen mecanismes flexibles realitzats en una única peça i que equivalen a mecanismes d'urpes realitzats amb diverses peces rígides unides per articulacions. Mitjançant el sinteritzat per làser, es fabriquen actuadors pneumàtics, amb diversos tipus de moviment, basats en la flexibilitat del material emprat en la seva fabricació. En conjunt es simplifiquen els sistemes arribant a realitzar urpes flexibles de diversos dits fabricades en una única peça. Per a avaluar la qualitat i frescor dels productes agroalimentaris durant la manipulació s'empren acceleròmetres localitzats en els dits de diverses urpes. Els acceleròmetres són econòmics i es comporten com a sensors tàctils intrínsecs, sense estar en contacte directe amb el producte evitant desgastos per aquest motiu, i permeten identificar les diferents fases d'prensió. Per aconseguir això es desenvolupa un procés específic del robot amb l'urpa, que palpa diverses vegades el producte. Es fabriquen diversos tipus d'urpa amb diferents tecnologies de mecanismes infra-actuats i sistemes jamming i es programa un algoritme original de processat de senyal que, amb diverses tècniques, és capaç d'extreure paràmetres dels acceleròmetres que serveixen per a avaluar la qualitat dels productes. Aquests paràmetres són correlacionats amb les dades d'assajos destructius que són habitualment emprats com a referència. Les millors capacitats s'aconsegueixen emprant urpes amb jamming assolint-se coeficients de correlació de 0,937 en índexs de maduresa amb mangos i 0,872 en fermesa d'albergínies. / Blanes Campos, C. (2016). Garras con sensores táctiles intrínsecos para manipular alimentos con robots [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/68481

This thesis provides a systematic guideline for evaluating, integrating and designing the collaborative end effector system associated with commercially available collaborative robots (cobots). On the basis of ISO/TS 15066:2016, customers’ demands are categorised as a macro demand framework and micro demand variables, which are parameterised and hierarchised. By analysing these parameters, the collaborative degree is determined to correlate the hazards with protection measures, which can demonstrate the collaborative safety level transparently. After that, the safety protection measures are specifically proposed corresponding to the potential risks. Finally, an industrial application is evaluated, a novel adapter utilised on the UR5 cobot is illustrated, and its characteristics for the collaborative operation are described.:1 Introduction 2 Robotic end effector in Assembly and Handling 3 Concept and Definition of Collaboration 4 Statement of the academic purpose and methodology 5 Guideline and Methodology for designing the Active Lightweight End-effector (ALE) for collaborative robots 6 Performance demand d4): multiplicity and versatility of end effector 7 A safety monitoring end effector (Adapter) for collaborative operations 8 A practical study of collaborative integration based on collaborative degree 9 Conclusion Appendix / Diese Arbeit liefert einen systematischen Leitfaden für die Bewertung, Integration und Gestaltung des kollaborierenden Endeffektorsystems in Verbindung mit kommerziell verfügbaren kollaborativen Robotern (Kobots). Auf der Grundlage der ISO/TS 15066:2016 werden die Anforderungen der Kunden in einen Makro-Bedarfsrahmen und Mikro-Bedarfsvariablen kategorisiert, die parametrisiert und hierarchisiert werden. Durch die Analyse dieser Parameter wird der Kollaborationsgrad ermittelt, um die Gefährdungen mit Schutzmaßnahmen zu korrelieren, die die kollaborierende Sicherheitsstufe transparent darstellen können. Danach werden die Schutzmaßnahmen entsprechend den potenziellen Risiken vorgeschlagen. Schließlich wird ein industrieller Anwendungsfall bewertet, ein neuartiger Adapter für den Einsatz an einem Roboter UR5 erläutert und dessen Kenndaten für den kollaborierenden Betrieb beschrieben.:1 Introduction 2 Robotic end effector in Assembly and Handling 3 Concept and Definition of Collaboration 4 Statement of the academic purpose and methodology 5 Guideline and Methodology for designing the Active Lightweight End-effector (ALE) for collaborative robots 6 Performance demand d4): multiplicity and versatility of end effector 7 A safety monitoring end effector (Adapter) for collaborative operations 8 A practical study of collaborative integration based on collaborative degree 9 Conclusion Appendix