Tag: 18MichaelE

Today was a rather light work day since I had already finished most of what Shihao had wanted me to do over the course of the two weeks.  In the first part of the day I worked on adjusting the starting position of the robot.  I used a computer program to adjust the angle of the servos so the robot would stand up straight.

After that I played around with the controller to see what kinds of motions the robot can do.  The final thing that Shihao had me work on was attaching sticky pads to the feet and hands, so the robot will get more traction.  This allowed it to complete some of its actions better because it could get a grip on the ground and not slip.

The past two weeks have been a lot of fun, and I have learned a lot about robotics, engineering, and college life.  It was a great experience and I recommend for all future juniors to participate in this program.

 

Today I worked on fine-tuning the code for the joystick to enable us to do the motions we want with the robot.  I also did some modifications to the legs of the robot.  Shihao wanted the legs to be able to bend backwards, but the original design only allowed the legs to bend forward or perfectly straight.  To do this, I had to remove the support system at the thigh and reattach it in a different way.  I used the bars that were in to original design, and attached it on the outside of the leg.  This allowed the leg to bend backwards because the support system would no longer hit the motor.  I am not quite finished with this because we need longer screws, but I will be able to finish this tomorrow.

I also got a chance to sit in on the weekly meeting with the professor.  All of the graduate students in the Duke robotics lab come together and talk about what they accomplished the last week.  This week’s meeting was a little different since it was the first meeting of the summer semester.  Dr. Hauser (the professor) and the students talked about goals they should accomplish over the summer.  Dr. Hauser also talked about some tournaments they could enter in.  One they are particularly interested in is a tournament where the teams have robots that need to preform task and be controlled by the students who are miles away.  The tournament is in October of 2019.  The prize money for first place is 8 million dollars, and for getting to certain stages, you get $200,000.  This is a high risk tournament to enter in though because Dr. Hauser believes the robot will cost $250,000 to build.

Today was a very successful day.  I was able to get the PS2 controller to hook up to the robot, and control the robot with it.  The robot can to a range of motions from walking to boxing to doing pushups to dancing.  It is actually quite funny to watch the robot do some of these actions.  Something I found interesting was the three different robots I have worked on were all from China and could all play soccer and box.  This makes me wonder if like robosoccer and roboboxing is a thing in China; It would be kind of funny to watch.  I also had to do some coding today as Shihao wanted me to customize one of the buttons.  He wanted to be able to press the “start” button and have the robot go back to its original starting position (arms stretched out horizontal to the ground while standing straight up).  To do this, I had to take some existing code and modify it.  Despite having limited programming knowledge, I was also successful with this task.

Today was also a very tense time in the lab because it was the monthly meeting with the professor.  Every PhD student, including Shihao, had to meet with the professor to give him an update on their work.  No one liked this meeting though because it is where they all get told their work is bad and they need to do more.

Today, I worked on getting the PS2 controller connected to the robot.  I had to first upload the code to the robot that allows it to be controlled by the PS2 controller.  After this I then started testing it and realized only two of the controller functions were working.  I tried re-uploading the code and even tried to change the code, but none of that worked.  I then tried to use a different set of code the manufacturer had provided, but the battery to the robot died.  Since the battery takes about 5-6 hours to charge, that was the end of connecting the controller for the day (thankfully because I was getting quite frustrated after trying to fix the controller problem for 4 hours).

After that, I helped Shihao with taking more measurements of the robot for the UDRF file.  We divided the robot into rectangular sections in order to get accurate measurements for the file.  I had the measure from the point of rotation for each motor to the center of mass of the rectangle the motor was in.  I also made some slight adjustments in the CAD file of the hands to fix flaws in the original design.  I had to widen the hand so it fit on the robot and thicken the part that connects to robot so it doesn’t break.

My day had an interesting start.  I saw a Lamborghini Aventador in the parking garage.

I started off the day by recalibrating the servo motors on Shihao’s robot, and I also finished installing the driver.  I also had to once again, remake the arms of the robot back to there initial position in order for the programs to work.  These task took about 2 hours for me complete because they were a little bit challenging and complex.

Shihao and I then went to the Duke’s Coinnovation lab to get our 3D printed hands.  Once we got there, we could not find our hands, so we asked a person working there and found out there were 3 Coinnovation labs.  Of course we realized that one hand had printed at one lab, and the other hand had printed at the other lab.  Shihao and I then had an hour of taking the buses around Duke’s large campus and walking in the heat to get our hands.  Overall, I have really enjoyed my first week working at Duke.

I started the day by working on uploading the driver to Shihao’s robot and calibrating the servos and the robot’s movements.  This did not go so well as I was having trouble getting the robot to move.  After about an hour of diagnosis, Shihao and I think that there is a problem with the driver itself.  Since the driver is software provided by a company in China, Shihao emailed them, and we hope to get a response tomorrow.

Next Shihao had me disassemble the robot into different sections: the forearms, arms, thighs, knees, legs, feet, and body.  Then I had to weight each of those sections and measure the length, width, and height of each piece.   This was a long task, made even longer by the fact that I put the robot back together the wrong way.  The measurements I took today I will use in the coming days to make a udrf file.  This allows me to create a virtual robot that has accurate size and weight to the real thing, so Shihao can do test with it.   I will also use that data to calculate the moment of inertia (I really hope Shihao helps me with this because I have no clue how to do that).

This morning, I began my day by making a CAD model of a hand for Shihao’s robot.  The new hand I made was in a box shape because Shihao wanted it to be a plane to contact the wall with so it has more stability.  The original hand was shaped like a human hand and had only one point of contact on the wall so it was not very stable.  The CAD model itself was fairly easy to make because Shihao gave me the files with the piece I needed to modify into the rectangular hand.  There was one problem, there were 200 different templates for pieces, and all the files were in Chinese.  This meant I had to open every file to find the one I wanted.  Of course the one I wanted was at the end so I spent a solid 30 minutes opening and converting files from a Solidworks file to a Fusion 360 file to see if it was the piece I needed.  After I finished the CAD model, Shihao sent it to the Duke innovation lab to be 3D printed.

The second thing I worked on was rebuilding the old robot for Shihao.  I had to take it all apart, remove all the extra piece and electronics he added, and then reassemble it.  This was a long and kind of boring task (I can see why he had me do it), but it payed off in the end.  When I finished it, Shihao connected it to his phone and let me play around with the motions it can do.  It can dance, do push ups and sit ups, punch, kick, and even dive as if to be a goalie in a soccer game.  All in all, it was another fun day working in the Duke robotics lab.

Today was my first day working with Dr. Hauser and PhD student, Shihao Wang at Duke University.  Surprisingly I was able to navigate my way from the Bryan Center parking garage to North Hall without getting lost, but I am pretty sure I took the long way there.  I met Shihao in his office, and he gave me a tour of the Duke robotics lab.  In the lab, they were working on two different robots.  One robot is for use in hospitals.  It is to be used to treat patiences with very contagious diseases such as Ebola.  The robot is operated by a person outside of the room so that no one else has to be in the room with the sick person.  The students are currently working on giving the robot the ability to remove its own hazmat suit so that robot itself does not become contaminated with the disease.  The second robot they are working on is a disaster relief robot.  This robot is being designed to go into an area that has experienced a natural disaster and go places where it is too dangerous for a human to go.  This robot is built, but does not yet function.

After exploring the lab, I went straight to work helping Shihao with a project of his own.  He is working on a robot, only about 2 feet tall and in a humanoid form, that can catch itself with its arms when pushed over.  The arms were not working quite right, so he had me remake them so that they would be able to stop the robot from falling.  To do so, I had to take the existing arms apart and remake them the way Shihao wanted them.  I also had to take out the battery and replace it since it had gone bad.  Once I fixed that robot, I went back to Shihao, and he gave me another task.  In the coming days, I have to take apart another one of his robots which he modified from a robot he bought.  He then wants me to rebuild it back to the original robot he bought so he can return it.  Overall, it was a fun first day working at Duke.