The thesis is equivalent to 20 weeks of full time studies.

This thesis is submitted to the Faculty of Engineering at Blekinge Institute of Technology in par- tial fulfilment of the requirements for the degree of Master of Science in Mechanical Engineering.

The thesis is equivalent to 20 weeks of full time studies.

The author declare that he are the sole author of this thesis and that he has not used any sources other than those listed in the bibliography and identified as references. He further declares that he have not submitted this thesis at any other institution to obtain a degree.

Contact Information:

Author: Nils Söderström

E-mail: nils.martin.soderstrom@gmail.com

University advisor:

Senior lecturer Alessandro Bertoni Department of Mechanical Engineering

Faculty of Engineering Internet : www.bth.se

Blekinge Institute of Technology Phone : +46 455 38 50 00

SE–371 79 Karlskrona, Sweden Fax : +46 455 38 50 57

Abstract

With the ever-growing market of six-axis robots in the previous years, many different kinds of robots have been introduced into the market. A smaller group of so-called collaborative robots have during this time gotten increased popularity. One of the inconveniences with this type of smaller robot is the lack of internal pneumatic ca- pabilities, which leads to external cables and tubing. This can cause problems with the cables attaching into unintended things and coming loose which could result in production stops, machine failure or other potential damage. Another part of this is that the external cabling hinders the robots sixth axis of rotation. The need for air pressure is to supply the common pneumatic grippers that the robots often use to gain pick-and-place capabilities.

Cobotech Kalmar AB is a company based in Kalmar which specializes in robot integration with these collaborative robots. The purpose of this project is to to- gether develop a product concept that can minimize the external cabling needed on the robot while still allow full rotation in the sixth axis. On top of this, the product should have plug-and-play capabilities to decrease the installation time of a robot unit.

The method to develop the results is the participatory action research (PAR), with the five different steps that the method includes: problem approach, design/plan- ning, acquiring data, analysis and reflection.

The result of the thesis shows that it is possible to decrease the external air tubes and installation time of a collaborative robot. This can be achieved while not decreasing the robots range of motion. By developing an innovative end-effector for the robot the main problems caused by pneumatic grippers can be eliminated.

The conclusion of this thesis is a concept product that has one air input and five outputs. This allows for the minimization of the external air tubes needed to only one. The product has a swivel function incorporated which allows full rotation of the robots sixth axis. This together with having the valves seated in the end-effector makes this conceptual product plug-and-play.

Keywords: Participatory action research, Collaborative robot, Robot in-

tegration, Product development.

Sammanfattning

Med en ständigt växande marknad för sexaxliga robotar de senaste åren har många olika robotar introducerats till marknaden. Av dessa har mindre kollaborativa rob- otar vuxit i popularitet. Ett problem med många av dessa mindre robotar är att de inte har tryckluft inbyggt i armen som många av sina större bröder vilket leder till att tryckluftslangar måste dras externt på roboten istället. Detta kan leda till att kabeln fastnar och rycks loss vilket i sin tur kan leda till produktionsstopp, att maskiner skadas eller annan potentiell skada på antingen maskiner eller människor.

Ett annat problem med att externt dra kabel är att den hindrar robotens sjätte ro- tationsaxel. Anledningen till att ha tryckluft på roboten är för att ha möjlighet att installera pneumatiska gripdon.

Cobotech Kalmar AB är ett företag bosatt i Kalmar som specialiserar sig på att in- tegrera kollaborativa robotar i olika tillverkande industrier. Syftet med detta projekt är att tillsammans utveckla ett produktkoncept som minimerar de externa kablarna på roboten och ser till att roboten får full rörlighet i sin sjätte rotationsaxel. Utöver detta ska produkten arbeta mot att ha plug-and-play kapacitet för att minska instal- lationstiden ute hos kunden.

Metoden som används för att få fram resultaten var deltagande aktionsforskning, som innehåller de följande fem stegen: problemformulering, design/planering, in- samling av data och reflektion.

Resultatet av denna rapport visar att det är möjligt att minska externa luftslangar samt mindksa installationstid på kollaborativa robotar. Detta kan göras utan att minska robotens funktion. Genom att utveckla en innovativ end-effector till Univer- sal robots produktutbud kan man eliminera de främsta problemen med pneumatiska gripdon.

Projektet har resulterat i en nära färdig konceptuell produkt som innovativt använder magnetiska miniatyrventiler tillsammans med en svivelfunktion för att enbart ha en tryckluftkabel som input till produkten. Produktens swivel-funktion frigör robotens sjätte rotations-axel. Detta tillsammans med ventilerna som sitter i produkten gör att den får plug-and-play funktion.

Nykelord: Deltagande aktionsforskning, Kollaborativ robot, Robot inte- gration, Produktutveckling

iii

Acknowledgments

In this part of the report, I would like to thank the persons that have helped this project forward.

Linus Nyman at Cobotech Kalmar AB for mentoring and help during this whole project. The discussions and help with the experiment was essential for this project.

Alessandro Bertoni at Blekinge tekniska högskola for help with the report struc- ture and method choice.

Several other people have helped as well. Thank you all.

v

Contents

Abstract i

Sammanfattning iii

Acknowledgments v

1 Introduction 1

1.1 Problem . . . . 2

1.1.1 Scope . . . . 3

1.2 Thesis question . . . . 3

1.3 Limitations . . . . 3

2 Background 5 2.1 UR Robots . . . . 5

2.2 Pneumatic Gripper . . . . 6

2.3 Solenoid valve . . . . 6

2.4 IP Code . . . . 6

2.5 Swivel . . . . 7

2.6 Participatory Action Research . . . . 8

3 Method 9 3.0.1 Problem approach . . . . 9

3.0.1.1 Understanding the problem . . . . 9

3.0.1.2 Market study . . . . 9

3.0.2 Design and planning . . . 10

3.0.2.1 Product needs . . . 10

3.0.2.2 Product specifications . . . 10

3.0.2.3 Concept generation . . . 12

3.0.2.4 Flow experiment . . . 12

3.0.2.5 Experiment conditions . . . 13

3.0.2.6 Equipment and setup . . . 13

3.0.2.7 Execution . . . 16

3.0.3 Acquiring data . . . 16

3.0.4 Analysis . . . 17

3.0.4.1 System development . . . 17

3.0.4.2 Technical models . . . 18

3.0.4.3 Product cost . . . 18

3.0.5 Reflection . . . 19

vii

4 Results 21

4.1 Problem approach . . . 21

4.1.1 Understanding the problem . . . 21

4.1.2 Market study . . . 21

4.1.3 Development of needs . . . 24

4.1.3.1 Interview . . . 25

4.1.3.2 Observation . . . 25

4.1.3.3 Internet research . . . 25

4.1.3.4 Interpretation of data to customer needs . . . 25

4.1.4 Product specifications . . . 27

4.1.5 Concept generation . . . 28

4.1.5.1 Air . . . 28

4.1.5.2 Mounting to robot . . . 29

4.1.5.3 Mounting gripper . . . 30

4.2 Analysis . . . 31

4.2.1 System development . . . 31

4.2.2 Concept iteration 1 . . . 31

4.2.2.1 Subsystem 1 - frame . . . 33

4.2.2.2 Subsystem 2 - air system . . . 34

4.2.2.3 Subsystem 3 - rotating system . . . 36

4.2.2.4 Subsystem 4 - seal for industrial environment . . . . 36

4.2.2.5 Flow experiment . . . 37

4.2.3 Concept iteration 2 . . . 40

4.2.4 Concept iteration 3 . . . 42

4.2.5 Concept iteration 4 . . . 43

4.2.6 Detailed development . . . 45

4.2.6.1 Technical models . . . 45

4.2.6.2 Flow analysis . . . 48

4.2.7 Final product concept . . . 50

4.2.7.1 Final specifications . . . 53

4.2.7.2 Summary of function . . . 54

4.2.7.3 Example images . . . 55

5 Discussion and reflection 57 5.0.1 General reflection . . . 57

5.0.2 Method reflection . . . 57

5.0.3 Result reflection . . . 58

5.0.3.1 Experiment reflection . . . 58

5.0.3.2 FEM analysis reflection . . . 58

5.0.4 Concluding reflection . . . 59

6 Conclusions 61 6.1 Contribution . . . 61

6.1.1 New knowledge . . . 61

6.1.2 New solutions . . . 62

7 Future Work 63

viii

References 65

A Supplemental Information 67

A.1 Appendix A . . . 67

A.1.1 Questions . . . 67

A.1.2 Questions and answers in English . . . 67

A.1.3 Questions and answers in Swedish . . . 69

A.2 Appendix B . . . 72

A.3 Appendix C . . . 73

A.4 Appendix D . . . 73

A.5 Appendix E . . . 75

A.6 Appendix F . . . 76

ix

List of Figures

1.1 Universal Robots robotic arm with a pneumatic vacuum end-effector

[32]. . . . 2

1.2 Universal Robots robotic arm with the rotation axis for each joint. . . 3

2.1 The shared gripper mount of Universal robot products. . . . 5

2.2 Description of swivel function. . . . 7

2.3 Steps in action research process [35]. . . . 8

3.1 An example of a concept matrix. . . 12

3.2 Schematics over the different experimental setups. . . 15

3.3 Setup of cylinder with valves. . . 15

4.1 RSP Swivel S5 [31]. . . 22

4.2 ZG rotary dispenser DVR40i4 [9]. . . 23

4.3 AP rotary dispenser [17]. . . 24

4.4 Air concept 1. . . 28

4.5 Air concept 2. . . 29

4.6 Air concept 3. . . 29

4.7 Mounting concept 1. . . 30

4.8 The Pugh matrix with the different brainstormed concepts evaluated. 31 4.9 Sketch over the concept. . . 32

4.10 The first CAD iteration of the concept. . . 33

4.11 Air path and location of seals. . . 35

4.12 Reference values from the compressor under different pressure. . . 37

4.13 Reference values from the Bürkert valve with input in port 1 and output in port 2. . . 38

4.14 Reference values from the Bürkert valve with input in port 2 and output in port 3. . . 38

4.15 The second CAD iteration of the concept. . . 41

4.16 The third CAD iteration of the concept. . . 43

4.17 The fourth CAD iteration of the concept. . . 44

4.18 Placement of loads and boundary conditions on the base plate. . . 46

4.19 Deflection (left) and von Mises stress (right) with a load of 300 N and with no holes in the base plate. . . 47

4.20 Deflection (left) and von Mises stress (right) with a load of 300 N and with holes in the base plate. . . 47

4.21 Deflection (left) and von Mises stress (right) with a load of 30000 N and with holes in the base plate. . . 47

4.22 Deflection (left) and von Mises stress (right) with a load of 30000 N axially applied and with holes in the base plate. . . 48

xi

4.23 Deflection (left) and von Mises stress (right) with a load of 15625 N

and with holes in the base plate. . . 48

4.24 The 3D model of the simulation. . . 49

4.25 The result of the first simulation with full flow. . . 49

4.26 The final product concept with all the internal parts. . . 52

4.27 The product without gripper mount. . . 55

4.28 The product mounted on a UR5e robot. . . 56

4.29 The product mounted on a UR5e without the gripper mount. . . 56

A.1 Mounting seat of the Bürkert valve. . . 73

A.2 Table of dimensions for mounting Turcon glyde ring in shaft [1] . . . 75

A.3 Table for the dimensions of the o-ring grove [29] . . . 76

xii

List of Tables

4.1 Target specification for the product. . . 27

4.2 Opening times for the SMC cylinder MGPM16-50Z. . . 39

4.3 The complete part list for the manufactured components. . . 51

4.4 The complete part list for the purchased components. . . 51

4.5 Final specification for the product. . . 54

xiii

Chapter 1

Introduction

This project is done by a student attending Blekinge tekniska högskola in cooperation with Cobotech Kalmar AB and in the following section the problem and purpose of this report will be described.

In today’s consumer society, speed is more important than ever. A customer does not want to wait several days or even weeks for a product bought online. This, in turn, increases the demand for manufacturing and shipping for companies and a natural response to this heightened demand is the implementation of automated robots.

To increase production and decrease costs, robots will have an obvious place in the future of the industry and an increased number of companies will have the opportu- nity to implement and use robots in their production. This implementation can, for example, lead to lower production cost, higher production stability and improving product quality [12].

The possible benefits of implementing robots in different production lines have made robotics a lucrative field and the increased production of robots can showcase that robots are here to stay [10].

With the increasing internet purchases in Sweden that takes place in all sectors with no seeming decline [16], robots will continue to gain market shares to ensure that the packaging and shipping for people’s shopping habits can be met.

There are several types of robots in the manufacturing industry today, but the robots mentioned in this report are always a six-axis robotic arm that together with different end-effectors are being used to pick up and place things accurately. This configuration makes the robot perfect for dull and dangerous jobs that humans often do not want to perform. There are numerous applications for this kind of robot but among the most common are: picking and placing parts, welding, painting, inserting screws and more. All these applications need different tools or end-effectors in com- bination with the robot to be preformed. End-effectors can be pneumatic grippers, drills or welding equipment. And in all groups, there are sub-groups to make sure the robot perform efficiently. For example, a smooth plain surface could need a vacuum gripper to pick it up while a rod might need a three-finger pneumatic gripper instead.

In later years, a subgroup of robots has emerged, so-called collaborative robots which can get integrated with humans in the manufacturing lines without the protective physical barriers like cages. These robots will stop their movement upon impact and therefore minimise the damage done to humans. This makes the robot more portable and can, therefore, be placed to suit the production needs. This combined with the versatile nature of the six-axis robotic arm makes it possible for most manufacturing industries or shipping companies to benefit from automation. The robot can be a servant to a CNC machine or inserting screws into different parts, as well as package parts into boxes and stack these boxes onto European pallets. Without physical

1

2 Chapter 1. Introduction barriers like cages, collaborative robots generate a more dynamic workplace for the whole company. Work that humans often see as dirty, dull or dangerous, a robot can perform effortlessly for a long time.

Cobotech Kalmar AB is a company located in Kalmar that focuses on robot in- tegration to automate manufacturing industry. The company develops the whole concept from the drawing table to programming of the robot and tailors the product to exactly match the needs of the customer. Cobotech specialise with 6-axis collab- orative universal robots (see figure 1.1).

Figure 1.1: Universal Robots robotic arm with a pneumatic vacuum end-effector [32].

1.1 Problem

As a general rule, pneumatic grippers are used on robots due to advantageous pric- ing, less complexity, higher gripping force and reliability compared to their electrical counterpart.

This generates a contradiction because many of the collaborative robots do not have

internal pneumatics integrated and therefore must have external tubing which can

hinder the robots 6th-axis rotation (see figure 1.2). External cables could possible

also get caught in things and break which causes a threat to the production stability

and the whole production chain. This is especially a problem when the robot uses

a double montage of grippers and possibly also a blowing function to remove debris

from the product, which is relatively common when serving CNC machines. A setup

like the one mentioned could potentially have up to five different pneumatic tubes

attached to the robot. Another problem of the pneumatic grippers is the slow instal-

lation time, to compare to the electrical counterpart which often has plug-and-play

capabilities and require little installation time. The pneumatic grippers require elec-

trical cables that link the robots computer and the valves. This generates downtime

and a hurdle to overcome before a pneumatic gripper is operational.

1.2. Thesis question 3

1.1.1 Scope

The scope of this thesis is to develop a product concept that minimizes the external cables on the robot. The cables are to be minimized while the robot still is fully functional with full use of the sixth axis. This report also aims to improve the in- stallation time of the robot by adding plug-and-play capabilities to the developed product. All these implementations should in no way make the robot perform worse in any aspect. The product should not hinder the robot in any way and have a weight that allows the robot to perform its intended work.

Figure 1.2: Universal Robots robotic arm with the rotation axis for each joint.

1.2 Thesis question

As of now, the robotic market does not offer any products that can fill the void of the problems stated in section 1.1. The challenge of this thesis is to develop the best possible solution to fill that void and therefore help the robot industry develop forward. This solution can be achieved by answering the following two thesis questions:

• How can a product combine the two features of minimizing external tubing and allow full rotation in the sixth axis?

• How can a product like this receive plug-and-play capabilities?

1.3 Limitations

The limitations of the thesis are that all developing work is done is with collaborative

robots in mind. Specifically Universal robots product UR16e will be the reference

robot to which the product is developed. This UR16e is the biggest robot in the

4 Chapter 1. Introduction product segment and will, therefore, generate the biggest load on the product. Other robots have different sets of benefits and downsides that are not covered in this report and the product could therefore not suit all different robots. This is done as Cobotech Kalmar AB uses this kind of robot primarily.

The product is made for industrial environment and is therefore possible not a perfect

fit for other environments. The product developed in this report is a conceptual

product and require more development before finished. The development process

with methods like brainstorming is highly individual and if the method of this thesis

were to be duplicated the results could differ from what has been described in this

thesis.

Chapter 2

Background

This part is to introduce knowledge about the different components that are used in the project. How they work will have importance when developing a product that uses many different components which all have to function together.

2.1 UR Robots

The robots that are being used primarily in Cobotech Kalmar AB are purchased from the manufacturer Universal robots. This manufacturer has a broad product segment but the focus will lay on the four different collaborate robots that are used by Cobotech. These four robots differ in shape and attributes. But the things that are important for this report are: how much the different robots can lift or what the payload of the robot is, and how to fasten a tool at the end of the robot. The four robots payload are up to: 3, 5, 10 and 16 Kg. The higher the payload the bigger the robot. The fastening geometry can be seen in figure 2.1 and are shared throughout all four robot models [18].

Figure 2.1: The shared gripper mount of Universal robot products.

5

6 Chapter 2. Background

2.2 Pneumatic Gripper

A pneumatic gripper is a tool used by robots to be able to grip different kinds of objects. Numerous companies manufacture various kind of grippers, all with their advantages and benefits. Three main types of grippers rely on pneumatics to func- tion, pneumatic grippers, magnetic grippers which have a similar way of functioning and the vacuum grippers which operates differently.

The regular pneumatic gripper which is the most common, open and closes its grip due to a piston in a cylinder that is being actuated by air. This is then reversed by applying air in another input and this will push the piston back in its initial position which will make the gripper release [14].

The next type of gripper is a magnetic gripper which also is operated with pneumat- ics. By actuating a cylinder the magnet will be pushed down and it will magnetize ferrous materials, "gripping" them. By then actuating the piston up to its initial position the magnetism will loosen och the "gripped" material will release from the gripper.

The vacuum gripper is a two-component system with a vacuum ejector that creates the vacuum that then at least one suction cup can utilize to lift a product.

The magnetic and pneumatic gripper use two air tubes to function while the vacuum gripper uses only one.

2.3 Solenoid valve

A solenoid valve is an electro-mechanical valve that is used to control fluids like air or water. The valve functions by having two positions, fully opened or fully closed.

These positions either allow or prevent flow through the valve. These two positions are achieved by energising a coil that by magnetism then lowers or raises a plunger from the valve orifice.

There are several different variants of solenoid valves but in this report, the 2/2 and 3/2 variants are of interest. The numbers stand for ports/positions. This means that the 2/2 variant has two ports (inlet and outlet) and two positions of the plunger.

This valve will either allow fluid through the valve or not. The 3/2 variant has three ports, inlet, outlet and exhaust port and the same two positions of the plunger as the 2/2 configuration. This configuration will allow the exhaust to flow through the exhaust port even when the valve is closed. This makes the 3/2 valves to be well suited for use with pneumatic actuators. The air that is inside the cylinder must move somewhere when the piston is actuated. If a 2/2 valve would have been used in this situation the air could not escape from the cylinder since the valve is closed, but by using the 3/2 valve the air inside the cylinder can be exhausted and the piston can return even when the valve is in the closed position [3].

2.4 IP Code

IP code stands for ingress protection code and classifies and rates the degree of how

well a casing or product withstands intrusion of dust, fluid or accidental contact.

2.5. Swivel 7 This is done by exposing a product for different degrees of fluid or dust to experi- mentally determine how well the product withstands these conditions. After testing the product receives a classification like IP54. The first number, the five in our case, tells how well the product withstands solid particles with a scale of 0 to 6. Zero means no resistance and six are dust tight with no ingress of particles. The second number, in our case four, tells how well the product is resistant to liquid ingress with a scale from 0 to 9K. Zero is again no resistance while 9K classification is given if the product withstands powerful, high-temperature water jets.

This classification makes it possible to get more detailed information about a product than just the label "waterproof" which tells very little. The IP54 would translate to a product that is dust protected and will withstand splashing of water from all sides [5].

2.5 Swivel

A swivel is in its most basic form a rod or a mount that is allowed to rotate axially free from the support structure. Synonyms for this function are rotary swivel, rotary union and more. This type of function can be found in many different applications and complexities, from swivel fittings which are built to minimize hose twisting. To rotary unions which allow excavators to rotate unlimited around its own axis with high pressure oil hoses without twisting and damaging the hoses.

This can be achieved by transferring the fluid from the support structure to the rotation part by utilizing the room between the two parts as a "reservoir". This enables the support structure to fill the "reservoir" independent from the rotating structure which then uses the fluid (see figure 2.2) and allows for continuous rotation [6].

Figure 2.2: Description of swivel function.

8 Chapter 2. Background

2.6 Participatory Action Research

Participatory action research differs from classical experimental research, because the researcher or his or her colleagues are involved and active members within the project that is being researched on. The participatory action research makes it possible to both pursue the truth about a subject and solve concrete problems simultaneously [34].

In classical experimental research, the researcher observes practitioners only when data is being collected. When this stage is done the researcher then takes dis- tance from the practitioners to develop knowledge. The participatory action research method instead allows the practitioners to be immersed in the research process to- gether with the researcher to jointly develop and learn from each other and the research being performed. This will allow the practitioners to be part of the devel- opment of the knowledge and then continue to use this knowledge even after the researcher is finished. When the research is done the researcher has generated new knowledge and has been part of developing new solutions (see figure 2.3).

Figure 2.3: Steps in action research process [35].

This way of researching allows for implementation and research simultaneously and to research valid problems together with the people who are affected by the prob- lem. The people affected are the ones with the best knowledge about the problem.

This also limits the possible methods due to the following criteria.

• The method needs to supply data for action input while still being valid.

• The method needs to make participation possible.

The different steps to solve the stated problem will follow the structure seen in figure 2.3.

The choice to use the participatory action research was made since the author also

works at Cobotech Kalmar AB and could utilize the resources there together with

the company to develop new knowledge.

Chapter 3

Method

In the following section the method to solve this project is formulated. This section will also cover what approaches that have been used, how and why they have been used.

The different steps to solve the problem formulation will follow the structure seen in figure 2.3.

The choice to use the participatory action research was made since the author also works at Cobotech Kalmar AB and could utilize the resources there together with the company to develop new knowledge.

3.0.1 Problem approach

The goal with the problem approach is to develop an understanding of the problem at hand and identify problems with existing solutions if any exists. The problem approach part was to understand the problem and what causes it.

Once the problem was clarified the next step was to discover what types of existing or similar solutions are available at this moment and how these solutions solve the problem or not. If not, then the last step of the problem approach is to determine in what way the existing problems have shortcomings and how to develop a product without those shortcomings.

3.0.1.1 Understanding the problem

The method of understanding the problem was to discuss the problem together with a representative at Cobotech. This contact is the person that realised the initial problem and works daily with robots so we could together come to an understanding of the problem and the objectives of the project.

This was done to get a firm understanding of the problem and what causes it. By understanding the problem it gives a solid platform to then conduct further research on the problem.

3.0.1.2 Market study

The market study was conducted at an early stage of the project to see what possible solutions already existed. And if no solution was to solve the problem then the market study could find out what the shortcomings of the product were.

This process was done by first developing a requirement for the product that was to be studied. This is done to only compare suitable products with each other and screen away other products. Once the requirement was done the search commenced and the following key words were used for the result: Rotary dispenser, Small rotary dispenser , Swivel and Small swivel.

9

10 Chapter 3. Method The result was then compared and analysed to see how these solutions stood against the problem that was formulated in the previous section.

3.0.2 Design and planning

The design and planning phase of the method was done to develop a framework to take on the problem in a structured way that would ensure the success of the project.

Also to make sure that the problem solved was the problems formulated in section 1.2. The design part of this phase is to design a method that can organize interaction between both research and action.

This part of the method was done in several steps, with parts of the design taken from Ulrich and Eppinger’s book Produktutveckling, konstruktion och design [4].

3.0.2.1 Product needs

The first part of this design was to develop the product needs. From the part Understanding the problem, in section 3.0.1.1, the basic needs of the product was developed. This needed to be improved to make sure all the needs of the product was covered. This was done by interviewing a representative at Cobotech Kalmar AB , and by observing how the robot integration works at Cobotech Kalmar AB . On top of the interviewing and observing there was also my participating during the whole development process of the robot integration to understand what parts are needed for the integration to function as intended (see section 3.0.3, Acquiring data).

Once the product needs were met an organising process began. This was done to determine which of the needs was primary and secondary to the product. The pri- mary needs are needs that are general to the product while the secondary needs are attributes that together build the primary needs. These needs were then sorted hier- archically after which secondary needs that were put together building the primary needs.

The second part of this section were to determine the relative importance of each need. This was an important step because most probably can not all the prod- uct needs combine into one product. Some of these needs will be removed through different trade-offs like strength versus size. After determining the needs relative importance to each other it became more clear which needs are more important than others. The less important needs can then more easily be terminated to give place for more important needs.

By doing all these steps the needs of the product was clear, both what needs the product require and which of these is more important then others. This gives once again a solid platform to conduct further research and makes sure that the needs are the foundation which the product is built on.

3.0.2.2 Product specifications

A product specification is a list of attributes the product require to meet the needs

that were determent in the previous section. These needs will solve the initial prob-

11 lem formulation. Specifications are divided into two parts, the target specification and the final specifications. The target specification are specifications that are ideal attributes for the product which should be worked towards from the start of the project. The final specifications are the specifications that the actual product will have as a finished product after several trade-offs and compromises.

To get the product specification right the method from Ulrich and Eppinger will be performed for both the target and final specifications.

Target specification

From the product needs several attributes were determined. These were determined by asking how the needs could be fulfilled, the attributes will together describe the product capabilities. The way of determining the specification was to look at each of the product needs and see how these needs could be satisfied. This was done to get a measurable attribute that can be checked if the product has reached the specification. An example of this could be that there is a need for the product to be light, an attribute of this could be the total weight of the product and the unit is then kilogram.

These attributes got the same relative importance as the needs it was derived from, this to keep the sat relative importance throughout the process.

From this list of attributes, a benchmark was performed on the same products as researched in the market study in section 3.0.1.2. The benchmark was done by just insert the attributes of each product into the list of attributes. This was done to make sure that the attributes of the concept product that will be developed has bet- ter attributes than its competition. The last step of the target specifications was to determine the ideal and/or acceptable target specifications of the product concept.

By doing this step the best and worst values of each attribute are generated and the worst values of the product will still be better then the competition. This will make sure that the product can be sold.

Final specification

The final specifications was determined when the product concept was in its last concept iteration.

These specifications were set by developing technical and financial models of the concept product. From these models, the target specifications can be updated into the specifications the model have and trade-offs can be performed. An example of a trade-off could be strength versus small size and weight or higher tolerances versus cost. All those attributes are important but since they contradict each other not everything can be done.

The technical models used here was FEM-analysis, both static stress and deflection analysis and flow analysis. These models are described more in-depth in section 3.0.4.2.

The reason to use target and final specifications were to ensure that the product

needs were being worked towards at all time of the developing process.

12 Chapter 3. Method

3.0.2.3 Concept generation

With the target specification in mind, the concept generation could begin. This was performed by conduct brainstorming several times to generate possible concepts that could fulfil the product needs. The concepts generated are a mix of innova- tive thoughts together with proven methods of solving a problem. The choice of brainstorming was made due to the simplicity of the process and since this had been performed many times beforehand in various courses and therefore the method was familiar to me. The brainstorming was executed by the author and done after the usual criteria: all ideas were good ideas, quantity is important and no criticism of ideas is allowed. The brainstorming was performed on several different parts that the product needed to have and these parts were: air and rotation, mounting to robot and mounting gripper. The reason for dividing the different parts of the concept was to be able to mix and match different ideas and together get a suitable one.

From these ideas, the most promising got into the eliminating process. This elim- inating process was performed by generating a Pugh matrix (see figure 3.1). This is a matrix where the different concepts are rated depending on differently weighted criteria and against a similar standard product which is used as a reference value.

The different criteria were based on the attributes that were determined by the target specifications. This was done to make sure that the needs were attended throughout the project.

Figure 3.1: An example of a concept matrix.

From figure 3.1 it can be seen that the concept AAA got the highest score and will, therefore, receive further development.

The reason to use this kind of eliminating process is to give all the concepts an equal chance to be chosen depending on the concept attributes. And by using this elimi- nating process the most promising concept will receive further development.

3.0.2.4 Flow experiment

The flow experiment was conducted to determine how much flow of air is needed for certain functions the product needs and how a specific valve impact the airflow.

In this experiment, the goal is to see how long time it takes for the valve chosen to

close a cylinder i.e. how long it takes for the cylinder to get filled with air and how

much flow this valve can generate. Another part of the experiment is to determine

if the valve is suitable for the blow function needed on the product and if the valve

is suitable to use for vacuum grippers. Bigger valves have bigger orifice where air

13 travels through and can, therefore, generate more flow. This is where a need for balance between size and flow is needed and it was determined that the best way to gather this information was to conduct an experiment.

3.0.2.5 Experiment conditions

The experiment was meant to simulate the conditions of the product in use, which converts to inside use, the normal temperature of 20 C

and with the normal air pressure of 1 bar.

All the pressures values measured are measured under pressure.

The experiment was done in Cobotech Kalmar AB office in Kalmar.

3.0.2.6 Equipment and setup The following tools were used:

Flow switch - PFMB7201-C8-B

The flow switch is an instrument to measure the flow rate for air or nitrogen. It is capable of measuring the flow between 2 and 200 l/m with an accuracy of ±3% [23].

Compact guide cylinder - MGPM16-50Z

This is a cylinder [21] with a bore of 16 mm in diameter and a stroke of 50 mm, this generates a total volume for the cylinder of 10053 mm

.

Filter regulator AW20K-F02CE-B

The filter regulator allows the user to regulate how much pressure that the outgo- ing airflow will have [22]. This particular regulator allows pressure between 0 - 10 bar.

Gripper MPC 075

This is a gripper from Schunk is a gripper with 15 mm stroke and an undefined bore [19].

Pressure Relief 3 Port Valve

On-off valve to regulate air into the experiment [25].

Type 6164 - pneumatic 3/2 solenoid valve

Bürkert Type 6164 pneumatic 3/2 solenoid valve is the valve of choice to experiment on due to its suitable specifications [2]. This valve has an orifice of 0.8 and will handle a pressure range of 0 - 7.5 bar. It should have a flow rate of 16 l/m.

Fully sunken manifold

The manifold for the solenoid valve to keep the valve in place and secure its function, see appendix A.3 for the drawing.

Vacuum generators

Festo vacuum generator type VH-14H-T4-PQ3-VQ3-RO2-M [7].

14 Chapter 3. Method

SMC miniature regulator

An SMC miniature regulator of any kind used to limit the outflow of the cylinder to simulate the product.

Other

Several tubes with different diameters were used to simulate approximately the di- ameters used in the product.

Several nipples were used in the size of m5 and m3 to simulate approximately the nipples that will be used in the product.

Several tube converters were used to convert the tubing between the different nipple sizes.

A standard electrical compressor and reservoir was used to generate the airflow and pressure.

A standard electrical unit capable of generating 24v current were used to open and close the valve.

Setup

In figure 3.2 the different experiment setups is shown. The compressor and air reser- voir are one unit which via one 8 mm tube are connected with the quick exhaust valve, this allows simple disconnection of pressure and airflow. This valve is then connected to one pressure gauge via 8 mm tube, this allows to select different pres- sures outgoing from the pressure gauge. The first experiment was then conducted by attaching the flow switch and taking control values of airflow at different pres- sure. The vacuum generator was tested by connecting its inlet port to the pneumatic pressure gauge to see what kind of airflow it needed to fully function. This was then disconnected to test the Bürkert valve which was then connected via one 4 mm tube to measure control values over this valve as well. Both port 1 to 2 and 2 to 3 were tested.

Subsequently, the second part of the experiment began and the MGPM16-50Z pneu- matic cylinder ingoing airport was attached to the outgoing port of the valve via 4 mm tubing and the outgoing port was attached to the SMC miniature regulator to simulate a Bürkert valve as one outgoing valve as well. This must be done since the product will be one valve for closing and for opening the cylinder and this might affect the test result (see figure 3.3). After the cylinder, the Schunk gripper was attached in the same manner.

For the last part of the experiment, the Fiesto vacuum generator was attached to the

Bürkert valve instead of the gripper and the outgoing Bürkert valve was removed,

this since the vacuum generator only have in-going air and exhausts the outgoing

air.

15

Figure 3.2: Schematics over the different experimental setups.

Figure 3.3: Setup of cylinder with valves.

16 Chapter 3. Method

3.0.2.7 Execution

The experiment was divided into three different parts. The first part, controlling the gear and measuring in and output airflow of the valve, compressor and the vacuum generator. This part also included blow to see how much air the valve delivered and if it could be used as a blow function to the product as well.

The second part are where the cylinder was tested and filled with air to see how fast it could open, this gives information if the valve are usable to be put into the product.

The reason to use this cylinder are that it has an large volume compared with the grippers normally used. The normal grippers lie within the range of 300-5000 mm

[24], while the cylinder have an volume of 10053 mm

. So if the valve work for this cylinder, it will, therefore, work with regular grippers. A test was also made on an regular gripper to confirm this.

The tests will be filmed with a mobile camera that record with 30 frames per second and by counting the frames the time it takes for the cylinder to open can be calcu- lated. This will be more exact then trying to use a stopwatch when the length of the opening time is most likely less than one second.

The last part of the experiment was to test a vacuum generator together with the valve and see how it performed with this attachment. This part was to get an understanding of the performance of the valves and will, therefore, only be tested by trying to create vacuum strong enough to grip different items around the workshop.

The measurement data will be taken incrementally from 1 to 8 bar since this is normal working conditions for a collaborative robot.

3.0.3 Acquiring data

The goal of this section is to describe how the data acquisition was performed and why the following methods were used.

According to C.R. Kothari [11] there are two types of data: primary and secondary.

The primary data are data which are collected for the first time, compared with the secondary data which is gathered by someone else and already processed into knowledge [11]. There are specific advantages for collecting either of the two data types. The primary data is collected from the source by using direct communication and generates four primary ways of collecting this data, by using interviews, obser- vation, questionnaires and schedules. This project will use interviews, observation and internet search for data collection. This is due to the qualitative nature of the problem which best can be answered by people with knowledge of robotics and in- dustry settings. The internet search part will be used to gather information about the market competition, grippers and components needed.

Interview

The collection of data by use of the interview method is done by presenting a series of questions to the subject which then are answered. A way of conducting this type of data collection is the personal interview, which includes an interviewer which ini- tiated the interview and asks the questions to an interviewee generally face-to-face.

This type of collecting data has several advantages. It is not limited to the explicit

17 questions written by the person interviewing but can proceed to follow a direction that is useful to the subject. This is suitable for exploratory research but hindering if several interviews should be comparable, which are not the case in this project.

Other advantageous properties of the interview method are that more information and a greater depth of information can be gathered with more flexibility compared to a questionnaire due to the interview’s unrestricted nature. The interview can be found in appendix A.1

Observation

The observation method works by searching for information by ocular inspection of a product in use or a specific behaviour without interacting with the target of the inquiry. This generates several benefits as a data collecting method. The subjec- tive bias can be eliminated if the observation is correctly executed. The information gathered is not altered by intentions or attitudes. The observation process for this project was continuous during the whole project.

Internet search

The internet search was performed to gather information on different components that could be useful for the product. First, a larger search was conducted to ensure that products that will be needed were available, products in this category were for example valves that could control airflow. This was done by first performing wider searches with keyword like ’pneumatic valve’ to then go narrower when results were found. This was also done in several different stages of the project when other fac- tors made the previous result obsolete. The internet was used due to the massive amounts of information that can be found quick and easily.

3.0.4 Analysis

The analysis step of this method was where the design och planning process was being performed with the information gathered in the acquiring data phase. The two phases together were being analysed and actions were taken from the results of the analysis. This results in a spiralling and iterative process, the data are being analysed, and actions from this are performed which change the data which then needs to be analysed again. This is done iteratively until a solution to the problem is found.

3.0.4.1 System development

The goal for this part was to use the acquired data from section 3.0.3 and combine it with the concept that won the eliminating process. This combination of concept and data could then be used to develop the different systems that would make the concept product is feasible.

For this, the first thing was to draw a quick sketch of how the product concept would function; which probable sub-system that the concept would require to function.

From this, the dimensions of this sketch could be estimated and a CAD model could be developed to make sure all the components work together.

Once this first test was performed the actual sub-systems was developed and the

18 Chapter 3. Method different components needed for the system to work identified. This identification was done by discussing how the system could work with a representative at Cobotech Kalmar AB and what components the system would require to work as intended.

These components were then researched to match the target specifications and then added into the product. From this, the iterative process began. This was done by identifying the less useful part of each new iteration and work on this to improve the concept while adding more and more detail into the CAD model. The less useful part of each iteration was identified by discussing with the contact at Cobotech Kalmar AB , by evaluating the iteration against the target specifications and by ocular control of the concept.

After four iterations of improvement, the concept was ready for the last iteration.

The last iteration was synonymous with the detailed development of the concept product. This was where all the final details that were too time-consuming to add to each iteration were added. This step continued to develop the technical models to see how the model perform in simulations and if any changes need to be done before the final specification can be determined.

By using this iterative way of developing the product got gradual improvements, the big problems with each iteration got identified first. With each iteration, the problems become smaller and smaller until each iteration has diminishing returns and there is no need for further iterations.

3.0.4.2 Technical models

The technical models include FEM simulation of the different parts made for this project and a compilation of all the parts to make sure that all parts fit together.

When working with simulations and CAD models they often come hand in hand, first, the CAD model is built by estimating the parameters of the part. This model is then remade or exported to the FEM analysis software to check how well the part performs with the estimated parameters. The result of the simulation will then make it possible to make an informed decision on if the part needs to be altered and in that case how it should be altered. This is however always a trade-off; a larger part will perform better to stress but is more expensive and heavier which are attributes that are bad for a component.

The technical models done on the product included one FEM-analysis to ensure that the model can handle the static and dynamic force that it will get exposed to. It will also include a flow analysis to ensure that the valves will get sufficient airflow similar to the experimental test. This is done to make sure that the model can not break under the loads and circumstances that it will encounter under its lifetime.

3.0.4.3 Product cost

The cost model will be made to make sure that the product can be produced within the budget and therefore will make a profit to the company. This will be made by taking price specifications on each part and then adding this together to a total sum.

Many of the standard components will have pricing set online and the parts that

are being made specifically for this product will be estimated by a representative

at Cobotech Kalmar AB . From this, the final specification could be assembled and

19 compared to the target specifications.

3.0.5 Reflection

The final part of the methodology was the reflection part. In this part, the validity

of the actions needs to be discussed. Could the design be done in other ways, and

in that case why is not that way chosen? What was the benefits of the used design

and what would be changed to next time? How has the design influenced the work

done and what could have been done to stop the influence? These are some of the

questions that every researcher need to ask themselves in a project like this.

Chapter 4

Results

The result of each phase will be presented below, this result is produced by following the method presented in chapter 3.

4.1 Problem approach

The first step of the project is to get more familiar with the problem stated in the introductory section and what the different products on the market have for attributes, and if these attributes can provide a solution to the problem.

4.1.1 Understanding the problem

From the communication with my supervisor at Cobotech Kalmar AB the follow- ing objectives of the project was communicated. These objectives are preliminary objectives to the project and will be further developed in a later stage.

1. The product must be compatible with Universal robots.

2. The product must minimize the external cabling on the robot.

3. The product must be able to use a double or single montage of a variety of grippers from different manufactures.

4. The product must be able to handle industrial air pressure (0-8 bar).

5. The product must allow full rotation of the sixth robotic axis.

6. The product must be able to plug-and-play

Of this, the last item on the list needs improved specification, as of now the valves which control the grippers are attached into the air cabinet beside the robot. This must then be connected with electricity so the robot and the valves can communicate.

This procedure can be time-consuming. The plug-and-play part of the product means that the assembling of the product can be made in advance and when the product is mounted on the robot it should just has air tube and electrical cable plugged in and then be ready for programming. This is possible due to that UR robots can connect to the different end-effectors from the end of the robot.

4.1.2 Market study

The market study was done in an initial stage of the project process to get a clear understanding of the competition and how other companies solve the problem at hand.

21

22 Chapter 4. Results Several companies are producing a variety of end-effectors for many industrial robots.

Since the collaborative robots and especially UR robots are a relatively small segment of the market products that are specialised for this brand are limited. This can easily be bridged with an adapter and products not specifically made for UR will therefore also be considered. The real problem is that most products are made for conventional industrial robots (non-collaborate) which in turn makes it possible for the robot in many cases to have a higher payload. With this in mind, the competing products must weigh less than 2.5 kg to still be relevant for the UR robot applications.

The following companies has at least one product that meets this requirement:

• Robot system products

• Zimmer group

• Applied robotics Robot system products

Robot system products (RSP) [30] is a buyout company from ABB with 50 employees that has a broad product segment for industrial robots, everything from tool changers to grippers and several accessories like tool stands for the robot.

RSP has a product family of swivels which can handle weight from 0 - 350 kg, of these swivels the Swivel S5 and Swivel S20 are within the weight limit.

Figure 4.1: RSP Swivel S5 [31].

The biggest swivel that still is within the weight limit is the S20 which weighs 1 kg and has a tool weight capacity of 200 N, approx 20 kg. This product has a total of six pneumatic channels (six inputs and six outputs) and communication from the robot to grippers is done via an 8-pole connector.

The smaller brother of the S20 is the S5 (see figure 4.1) which weighs 0.5 kg and can

handle four pneumatic channels (four inputs and four outputs) and communication

from the robot to grippers is done via an 8-pole connector. The max tool weight

capacity is 50 N approx 5 kg.

4.1. Problem approach 23 Both of these swivels have a maximum of 150 l/m of airflow, have an operating range of 0 - 10 bar of air pressure and allow full rotation of the sixth axis. The IP classifi- cation of both swivels are IP 54.

Zimmer group

Zimmer Group (ZG) [8] is a company located in Germany which focuses on han- dling, linear, industrial and damping technology. The company today has over 1200 employees and a turnover of 165 million euros annually.

Zimmer group has a product family of rotary distributors where four meet the weight demand of 2.5 kg. The smallest member of the family is the DVR40I4 (see figure 4.2) which weighs in at 0.2 kg and has a tool handling weight of max 15 Kg. The tool has 4 pneumatic channels and a four-pole electrical energy transfer. The tool has an IP40 classification.

Figure 4.2: ZG rotary dispenser DVR40i4 [9].

The largest rotary dispenser from the ZG within the 2.5 kg limit is the DVR80I6 which weighs 2 kg, has tool handling weight of 50 kg and six pneumatic channels.

The rotary dispenser has a 6 pole electrical energy transfer and an IP classification of IP64. Between the smallest and largest, product Zimmer group has two incremen- tally larger products.

Applied robotics

Applied robotics (AP) [17] is a company based in Germany which specializes in end-of-arm tooling for automation processes. These tooling products include tool changers, collision sensors, grippers, and other end effectors.

AP has four rotary dispensers which fit within the weight limit with the smallest one being the PRD2-31.5 model. This model weighs 0.58 kg and has two pneumatic feed-throughs. It has a max handling weight of 6 kg and an electrical module for eight discrete inputs.

The largest model under 2.5 kg is the PRD2-63 which weighs in at 2.41 kg (see figure

4.3). This model also has two pneumatic feed-through, a handling weight of 35 kg

24 Chapter 4. Results and an electrical module for eight discrete inputs.

Figure 4.3: AP rotary dispenser [17].

Between the PRD2-31.5 model and PRD2-63 there are two incremental larger models. The three largest models can be ordered with four pneumatic channels.

Conclusion of market study

From this initial market study, one can quickly realise that most of the products pre- sented are intended for larger machines than the UR robots available. The products presented also have the same number of air channels in as out, which will, in turn, need as many cables as inputs for utilization of all channels. This is a problem for robots that need external air tubes as this would mean that either the robot would need a matching number of air tubes from the air supply to the swivel or one tube connected to a divider then connected to the different inputs in the swivel. Some of the products has a low IP class which means they could break in contact with cutting fluid or dirt ingress.

From this, it can be seen that none of these products fulfil all the preliminary ob- jectives sat in section 4.1.1. These products are not plug-and-play nor will they necessary reduce the number of external cables the robot would need to secure the function of these products.

4.1.3 Development of needs

To gain a full understanding of the product that was being developed further in-

formation needed to be gathered. This was done according to the three different

methods described in section 3.0.3.

4.1. Problem approach 25 4.1.3.1 Interview

The interview with my supervisor at Cobotech Kalmar AB was conducted in an early phase of the project with the purpose to gather general information about robot integration and the difficulties that come with the industry. The first conclusion from this interview was that the product will be subjected to a lot of cuttingfluid and must, therefore, have good sealing to hinder this fluid to penetrate the product.

The second conclusion from this interview was that most of the negative aspects of the pneumatic grippers can be bypassed if the product can be made plug-and-play and easily made to fit many different grippers. The interview also shed light on which grippers that are used and how this affect the dimensions of components.

The full interview can be found in A.1.

4.1.3.2 Observation

Several observations have been conducted by observing my supervisor while installing the grippers pneumatic components. These observations have enabled an under- standing of how the process for robot integration works and what this product could bring to the production industry.

The experience of my part-time job has granted over the last 18 months is substantial and will also be taken into account when producing the needs for this product.

4.1.3.3 Internet research

The internet research was initially conducted to search for different pneumatic grip- pers and what they needed to function. The result was that most grippers function between 0-8 bar of air pressure which are supplied by nipple between M3 and G1/4 in size. Conversion between the sizes are viable but might result in longer closing and opening times for the gripper due to the cylinder filling more slowly with air.

The grippers also have different hole patterns depending on size and manufacturer which could complicate a one-size-fit-all fastening solution. From this, the internet search became more about the different internal components of the product. Different valves, seals and other equipment was sought for.

4.1.3.4 Interpretation of data to customer needs

In this section, the data gathered from the different methods will be refined by the process described in section 3.0.2.1. By following this method the needs were generated quickly and the primary and secondary needs identified.

The primary needs are indicated with a dot and the secondary ones with a line. The relative importance of each need is determined by the number of asterisks (*) for each need, more asterisks equal higher importance with a maximum of three asterisks and a minimum of zero.

• The product is easily installed.

– (**)The product can easily be mounted on the robot.

– (***) The product does not require long setup time.

26 Chapter 4. Results

– (*)The product can easily be assembled.

– The product can easily be disassembled.

• The product is reliable.

– (*) The product can withstand industrial level wear.

– (***) The product can withstand cutting fluid.

– (***) The product can withstand dust.

– (**) The product can withstand collisions.

– (*) The product is rust-resistant.

– (***) The product maintains its precision for all its lifetime.

– (***) The product has the same tolerance for repeatability as the robot.

– (*) The product generates little friction.

• The product has a pleasing collaborative feeling about it.

– (***) The product is intuitive to use.

– (*)The product is quiet.

– (*) The product is comfortable to hold and grip.

– (*) The product is easily visible.

– (**) The product can be used to guide the robot.

– (***) The product is not dangerous.

• The product can rotate freely around its axis.

• The product minimizes air tubes needed for the grippers to work.

• The product can handle up to 8 bar of air pressure.

• The product improves the robot performance – (**) The product is compact.

– (*) The product can be used in cramped spaces.

– (*) The product is small.

– (***) The product does not hinder the robot’s motion.

– (**) The product allows a fast closing time of the grippers.

• The product can connect air to a pair of grippers plus one blower.

• The product allows the user to choose which gripper to activate.

• The product works with great flexibility and a range of robots and grippers used

– (***) The product fits all the UR robots.

4.1. Problem approach 27 – The product fits many different collaborative robots.

– (**) The product can communicate with the robot.

– (*) The product allows a broad spectrum of different grippers to be mounted.

– (**) The product allows for flexibility of how many grippers that can be mounted.

From these customer needs there is now a solid foundation to continue work on.

4.1.4 Product specifications

From the customer needs and the starting objective, the target specification was generated. This was done by studying the benchmarked products to see their take on the problem and then generating target specifications that better suited the needs for the product that was being developed.

Table 4.1: Target specification for the product.

Nr Measurable attribute Importance

(1-5) Target speci-

fication Unit 1 Time to setup the

product 4 max 300 s

2 Time to assemble/dis-

assemble the product 2 max 900 s

3 IP classification class 5 at least IP54 IPxx

4 Mounting Precision 4 at least h7 mm

5 Noise emission 2 max 60 db

6 Total weight 3 max 2.5 kg

7 Comfortable grip 2 - subjective

8 Air inputs 5 max 1 nr

9 Air outputs 3 exactly 5 nr

10 Operating pressure 5 between 0

and 8 Bar

11 Maximal diameter 4 max 63 mm

12 Gripper closing time 3 at least 1.5 s

13 Max axial force 3 exactly 150 N

14 Max bending moment 3 exactly 9 Nm

15 Airflow 5 max 200 l/min

16 Maximum height 5 max 60 mm

17 Handling weight 3 exactly 15 Kg

18 Voltage 3 exactly 24 V

19 Work temperature

range 2 between 5

and 60

C◦

20 Manufacturing cost 5 Max 10000 kr

28 Chapter 4. Results In appendix A.2 the benchmark with the other researched products can be found.

From this benchmark, it could be seen that none of the other products has plug-and- play capabilities and that the products would not decrease the amount of external tubing on the robot. All of the products could, however, free up the robots sixth rotation axis.

From this, it can be concluded that a new product needs to be constructed to better suit the needs.

4.1.5 Concept generation

This new product was divided into three bigger parts, air and rotation, mounting to the robot, and mounting gripper. The air part aimed to solve the problem of generating air pressure and airflow rate to be able to both close the grippers and generate pressure to hold the grip. This while the robot have free rotation around the sixth axis. The mounting to the robot part addresses the problem of mounting the product to the robot. And last, the mounting gripper part aims to solve the problem of mounting one or several different kinds of grippers to the product. This brainstorming generated several concepts where the best ones are presented.

4.1.5.1 Air

Tube leading to reservoir (A)

This concept work by having one air tube connected to the product from which the product uses the industry’s own air reservoir to access airflow and pressure. By having a kind of swivel mechanism allowing the rotation around the robots sixth axis. This means that one input of air is connected to the product as seen in figure 4.4, but this solution will guarantee that sufficient air is generated to the product.

Figure 4.4: Air concept 1.

Generator inside end-effector (B)

This concept would use a small air generator inside the product to generate air pres-

sure and airflow. This would not use any input air tubing and instead use electricity