2017 Team 7 – Automation of Handling Objects
PROJECT OVERVIEW
A local company came to the University of Maine with a project they needed help with. The company had recently purchased a used ABB FlexPicker Robot that they are hoping to re-purpose to meet their needs. Their hope is to get the robot to unpack objects from boxes and arrange them into a wash bin. The objects would then be sent down a conveyor belt and go through a wash process.
PHOTOS
To view our photo album for this project, click here.
WEEKLY UPDATE
Week 4/24
This week we cut the angle iron to make the tray platform, bent sheet metal to make the magazines, and are working on welding everything together. The cell is near completion, we just have a little more welding to complete and one more part to machine. Once we complete these last few tasks, we the cell should be finished. The end effector was delivered this week, which means we can machine the angle block to attach the end effector and the robot will be physically complete. All that will be left to do is wire and program the robot, which can be completed by the robotic expert as soon as he is available.
Weeks 4/10 and 4/17
This week our team ordered the end effector from McMaster, which we are hoping will arrive by next week at the latest. In the mean time, we have finished the cell design in Solid Works and have begun to machine the necessary parts. Some of the parts (such as the angle block to attach the end effector) can not be made until we receive the end effector. We have reached out to the robotic expert to schedule a time for him to help us finish the wiring, since we should have the cell completed by the end of next week.
Week 4/3
This week we heard back from our client about the end effector, and he liked the option but wanted to make sure that it could handle the acceleration (and still get 32 objects into the tray in one minute) of the robot and that we would be able to fit the last row of tiles in with it. To prove these two capabilities, we calculated out accelerations, and drafted the entire robot/cell in solid works.
The calculations showed that the end effector could only handle 1g (about 10m/s^2) of acceleration while holding the object, which is about a tenth of the maximum acceleration of the robot. Using this data and the distance the robot needs to travel for each step of one cycle, we determined that the robot could finish one cycle in just under 2 seconds. With this time to complete one cycle, loading 32 objects into the tray in approximately 63 seconds. The distances used were a worse case scenario, and therefore the robot and end effector should be able to meet the 32 per minute goal without surpassing the acceleration limits.
The robot, cell, magazines, tray, supports, and the end effector are almost completely drafted up. Using the drawings, our team has determined that if we change the angle of the objects to 80˚ instead of 60˚, we should be able to fit the last row of objects into the tray. Once the drawings are completed, we will send the drawing package and calculations to our client for review and hopefully they will agree that the end effector will meet the necessary parameters. Hopefully we can get the end effector ordered the beginning of this week so we can get the robotics expert to come in and help us get the robot completely up and running in the next couple weeks.
Week 3/20 and 3/27
Coming back from break, our team is working on getting orders made for the parts that we need to ensure they will arrive in time. The main thing that we have been researching to order is an end effector. We found an end effector that we believe will fulfill all of our needs and comes included with everything we need to get it running for much less money than other options. We sent the specs to our client, hoping they would find it to be as good of a option as we do.
In addition to end effector research, we also have been working a lot to get the magazines designed and cell set up so that we can have the robotics expert come back and help us wire and program the robot. We made a prototype magazine that seems to work well. Now we just need to design a way to attach the magazines to the robot cell, and the magazines should be complete. We also removed a post from the robot to better suit our needs. This post we will reattach in a different location to hold the tray once we have made the supports for the tray. Once we have the magazines and supports finished the cell should be complete, leaving just the end effector left to complete before we can tackle the wiring and programming.
Week 2/27
This week we met with the local robotic expert. He gave us a few new things to think about, and got the robot running full speed with a loop. This means we are able to do all the programming from the pendant, even for the purposes of our client, meaning the computer is not necessary, just a bonus. The expert suggested that we use two magazines instead of one, and just use positioning with the robot instead of having the tiles index up. This will make things a lot easier not only for us to create, but also for our client to load.
While the expert was meeting with us, we noticed that we may have some trouble fitting the last row of tiles into the tray due to clearance issues. The wall height of the tray make the last row difficult for the robot to reach, therefore it may be difficult to achieve the total number of objects in each wash tray that out client wants. This may be solved by adjusting the load angle of the objects, but we will have to do more research to be sure.
To wrap up our meeting with the expert, he said if we completed the cell and got the end effector, he would come back and help us finish wiring and programming the robot. There is still a lot of work to do, but having help with the wiring and programming is welcome news.
Week 2/13 and 2/20
During our last team meeting, we got the robot to run a small program (video can be found here). The programming was written entirely on the pendant (controller), not the computer. This works fine for our purposes of testing and writing the code, but we still need to get the computer program working so that our client can use the robot and run the program more efficiently. The robotic expert we contacted said he could come help us next week and seemed confident he could help us with both the input and output wiring and the programming troubles.
We have created a rough model of an end effector and a magazine system. The robot has a weight limit of approximately 1kg, which includes the weight of the end effector. Because of this weight issue, we decided it would not be feasible to add a fifth axis to the end effector (which would allow us to rotate the objects from horizontal, to a 60˚ angle in the bin). Instead we are going to have an end effector at a fixed angle, and load the objects into the robot at that angle. This saves weight and reduces complexity. To load the objects we are going to have a magazine system, that uses a linear actuator and a proximity sensor to ensure the objects are in the same place each time the robot goes to grab one. The rough model of these we are working on assembling onto the model of the cell and robot. Going forward we are going to meet with the robotic specialist to get the wiring and programming figured out, and we are going to get a prototype of the end effector and magazine system made so we can begin testing.
Week 1/30 and 2/6
This week our team took the conveyors, vision system, and panel off the robot. With what we are going to have the robot doing, these aren’t necessary. This was quite the process, but now that we have stripped all of the excess stuff, it is much easier to begin adding things to the robot to make it work like we need. We have generated a list of inputs and outputs that we need to wire for the robot, and have contacted the local robotic specialist to see if they can offer any help. Going forward we are going to design a prototype end effector, magazine system to index the objects up, and a way to hold the wash bin in the cell.
Week 1/16 and 1/23
Coming back from break, our group has a lot of work ahead of us, but a much clearer direction for our project. We have recently met with our client for the project and he advised us that the wiring and programming shouldn’t be our focus. Instead we should seek outside help for those portions of the project and focus on designing an end effector, a way to load and unload the objects from the robot, and figure out a way to angle the objects to load them into the wash bins (either by loading them at an angle, or rotating them with an end effector). This makes the project much more manageable for us since wiring and programming isn’t exactly our forte. Going forward we are creating a list of inputs and outputs we need to run the robot, brainstorming designs for end effectors, and determining the best and easiest way to load and unload the objects to integrate it with the existing work cell. Once we have a clear path and plan in place we will contact a local robotic company recommended to us and see if they can help us program and wire the robot to meet our client’s needs.
Week of 11/28 and 12/5
Our team has separated into two groups to expedite the process of getting this robot up an running. One group has been working on the wiring, the other has been working on the programming. The wiring group has gotten in contact with Ken Williams to possibly help complete the wiring. He seems confident that he could help with this project, but he informed us that we do not have any wiring diagrams for the control box that we need to wire. In order to finish the wiring so we can progress further into the project we will need these diagrams. We are currently in the process of contacting the company that made the work cell to see if we can get a copy of the wiring diagrams. The group working on the programming has determined that this robot does not have PLC, instead it uses a computer program to control the robot. We have been attempting to get the program “Robotware” up and running to begin getting this set up with the robot. We are still looking into PLC as it may be easier to add PLC than to get the computer program working, but that has yet to be determined.
Week of 11/7 and 11/14
We are currently trying to set up a time when someone from the AMC with an electrical wiring background can help us finish wiring the robot. We have hit a wall with the wiring due to poorly labeled schematics and wires. We are hoping that someone with more experience with these types of projects will be able to help us sort out the rest of the wiring, which needs to be completed before we can go any further with the robot. In addition to trying to finish the wiring, we are also looking into options for end effectors and how to best hold the tiles.
Week of 10/31
This week we met with our client to go over our progress and talk about the scope of the project. Instead of loading onto a conveyor, we will instead be unpacking a box of objects and will stack them into a wash bin. This wash bin will then go down a gravity conveyor and through the washing process. Because of this change we will need to do some research on how we want to pick up the objects with the robot. We also continued wiring the robot this week, but may need a little assistance in finishing the wiring due to the poorly labeled wiring.
Week of 10/24
This week we began wiring the robot. There is still a lot to be done, but we made some good progress on it. The wiring that needs to be done is reconnecting the control box to the robot itself. These were disconnected for transport, and need to be reconnected. We focused on the output wiring, between the robot and the control box, next week we will hopefully get the inputs done.
Week of 10/17
This week we cleaned the robot up and began sorting out the wiring. With the help of the AMC staff we managed to get the robot hooked up to the electricity, and installed the battery. Now that we have power going to the robot we can begin to wire and calibrate the robot for our needs. This is going to be a long process because of the many unlabeled and unknown wiring configurations, but after we get that sorted out things should be much easier to work on and figure out. This week we are also researching about PLC and ABB programming and trying to find resources to help us learn how to use these types of programming.
This week the robot was brought to the Advanced Manufacturing Center on campus for us to work on the robot more easily.
Week of 10/3
This week we met with out client and talked about the specifics of the project. The robot has all the parts and pieces, and we were able to get the robot to move, but it has a long way to go before it will be able to move on its own. The vision system needs to be set up, a new battery needs to be installed, a lot of programming and calibration will need to be done, as we will need to re-configure the existing conveyor belt system to better meet the client’s needs.
Weeks 9/5 – 9/26
Our team began researching information about PLCs (program logic controllers) and ABB FlexPicker robots to help get an idea of what this project will involve.
TEAM BIOGRAPHIES
Lukas Brewington
Lukas is a fourth year student at the University of Maine Orono pursuing his B.S. in Mechanical Engineering Technology. He spent three years as a CNC Machinist at Nichols Portland before college. He has been an intern at D&G Machine for the last two years and during the school year he works at the Advanced Manufacturing Center. His interests include mechanical design, prototyping and manufacturing. After graduation in 2017 he plans on working for D&G Machine as a Product Engineer.
Jordan Bleakney
Jordan is a fourth year student at the University of Maine at Orono. He is majoring in Mechanical Engineering Technology and minoring in Renewable Energy Engineering. He worked at Boss Snowplow this summer as a Design Engineer Intern. During the school year he works as a Teaching Assistant for the Machine Tool Lab at UMaine, and as a Shop Aid at Bangor Truck Equipment. After graduation in 2017, Jordan plans on working as an engineer for a local engineering firm/facility.
Alex Ouellette
Alex is from Kennebunk, Maine and in his senior year at the University of Maine as Mechanical Engineering Technology major. He transferred here after three years at Southern Maine Community College where he received an Associate’s degree in Precision Machining and certificate in Welding. He has worked at the Advanced Manufacturing Center on campus for over 2 years, and while at SMCC he worked at Don’s Sheet Metal and Soleras Advanced Coatings. His interests include 3D printing and mechanical design, and while at the University of Maine he has joined the 3D printing club.
Connor Morrison
Connor Morrison is from Biddeford Maine, in his senior year at the University of Maine enrolled in the Mechanical Engineering Technology program. He has spent his last two summers as a design engineer intern, and a process engineer intern at Rochester Electronics in Newburyport Massachusetts. His hobbies include weight lifting, playing piano, and skiing.
Caleb Rivera
Caleb Rivera is in his senior year at the University of Maine, finishing up a Bachelor’s Degree in Mechanical Engineering Technology. His hometown is Vassalboro, Maine. During the previous two summers he has worked with the Department of Transportation as a Transportation Aide. His future plans include pursuing a Master’s Degree in Business while working for a manufacturing and/or marketing engineering company. Hobbies include four wheeling, working out, traveling, and trying new foods.