Jacob Gardner - Individual Reflection

I’ve found Mech 250 to be one of my most interesting, fun, and time-consuming classes I’ve ever had. Coming up with unique strategies and designs, and seeing them come to life was an amazing thing to watch.

Like Brendon, I’m extremely happy with group. All members contributed the most they could. While some members were highly experienced with some of the engineering processes, others were not. Those who were not experienced did other tasks, enabling the team to work quickly and efficiently. However, the less experienced members weren’t exposed to some of the engineering processes, so that while the team got the work done, not everyone learned all the processes.

Prior to this class, the only manufacturing I had learned was from doing Robotics in high school. Only one of the two years I was in the class (we had a class too, not just a team) we participated in FIRST. However both years we participated in OCCRA (Oakland County Competitive Robotics Association) which, while a fun experience, didn’t teach us all that much about manufacturing. While we designed our robot with drawings and the like, we never used CAD during OCCRA and weren’t allowed to use precision machinery in order to keep the competition “fair.” I got used to that method of doing things; eye-balling almost everything, figuring out all the components as we went along. This class taught me that that method really isn’t the best way to do things, or even a good way. A finished machine should be precision-engineered. Planning and modeling should all be done prior to manufacturing to ensure everything fits right, works right, and will hold up. The real world of engineering doesn’t work like a high school Robotics team.

Our bot itself was quite an interesting piece of machinery; or “pieces” rather. Decepticon didn’t work quite as we had hoped, mostly due to traction issues caused by weight distribution not favoring the wheels enough. While we were never able to drive Decepticon into the opponents’ hole, it did perform its secondary function of distracting the other team quite well. Autobot performed admirably despite our decision to not use the roller due to questionable legality post-timed running and its sensitivity to the differing control boxes.

Overall I am happy we decided on the 2-bot strategy and went with it. While it may not have ended up working quite as well as hoped, it was a great learning experience. I learned how there are always going to be unique approaches to challenges, and sometimes they can work amazingly, and of course sometimes poorly, but that they’re often worth a shot.

The course could be improved mostly with clearer rules; the sandbagging for instance. We were told that sandbagging wasn’t allowed though we were later told it was allowed for the purpose of increasing traction, which is only really useful reason in this game. However it was too late at that point to change anything. Also, the rules on the “degrees of freedom” I feel should be changed. If one input is used to provide the desired output, even if the desired output involves multiple parts doing different (but not independent) things, it should still be considered a single degree of freedom even if it is done electronically; for instance, controlling a roller and a driving motor with the same joystick movement.


I could have improved my performance by making a greater effort to learn some of the engineering processes that I wasn’t as familiar with such as CAD and 3D printing. While I definitely learned about them, I didn’t learn as much as I would have liked, and not from lack of available information. I simply worked on different parts of the project.

Thank you to my teammates for working so hard and making this experience a really enjoyable one and for working so hard. And thank you to the ME250staff, for teaching me about the world of engineering and doing so in an interesting and enthusiastic way.

-Jacob Gardner

Final Summary

Autobot:





Decepticon:

Our final machines are shown above. Autobot, which is the offensive bot of the two, was what consumed the majority of our time. The frame consisted of an interior layer of 1/8^th inch aluminum and the outer layer was ¼” acrylic. This provided us with a sturdy, yet lightweight machine. Initially, we had planned for our bot to travel up the wave field in the back and get all of the balls off of the hill. Later on in the project, we determined that it would be more advantageous and less risky to alter our strategy and go for the front balls, and then either our opponents front balls or the balls at the base of the tower. This decision came from our bot being quicker than our calculations predicted as well as the seeding round being unopposed with all of the front balls open and low scores being posted in test runs by all of the teams. Embracing the team name we were assigned, Autobot has a drawing of a luchador rabbit etched into the side plates. Decepticon also has space invader on its body plate.

Our timing circuit was difficult, but eventually got made. We didn’t use the circuit for the competition because we feared that it may be triggered unintended current through the motors before the match started. After we placed 4^th in the competition we tried the circuit out and it worked exactly as planned and lasted about 80 seconds before shutting of the power and stopping the motion of the roller.

Our pulley’s and belt turned out better than expected. The 3D printed pulleys worked wonderfully and the belt provided great grip and comfort knowing that we couldn’t be pushed around or fall into our own hole when depositing balls.

Decepticon didn’t turn out as well as Autobot did, and would have benefitted from allowing more testing time. Decepticon had trouble moving due to lack of grip and weight on the wheels and loss of power due to wire length. If we had known that we could add weight solely for the purpose of increasing traction we definitely would have done so. We had a few problems with the size of Decepticon and ended creating a larger body for it the day before the seeding round. Other than those two problems, the 3D printed pillow block bearings and concept of Decepticon turned out as planned.

Jimmy Cofsky- Individual Reflection

I am not surprised that ME 250 is a required beginning class for those who think they plan on majoring in mechanical engineering. It is a grueling class that those who have a true desire to continue similar type work in their futures will enjoy and excel in. ME 250 not only tested us individually, but as a group.

I wasn’t sure what to expect when I entered the first day of class at 8:30 am half awake. For the first few weeks, the class remained pretty slow. The first few assignments were pretty straightforward and I was kind of excited to learn CAD for the first time. Then, we started to get our groups together and I am happy that I landed with Team Fluffy Bunny. One of the most valuable things learned in ME250 is how to work with a team to get a project completed. Throughout the semester we were able to work well together. We would try and work around our other hectic parts of our schedules to meet and get all of the assignments done. Our timing and awareness of what we needed to turn in progressed as the class went along. The assignments began to speed up and the idea of our machine construction and competition rapidly came into sight.

That is one of the things that caught me off guard the most in the class. The final project of having a complete machine that can run and score points in “the arena” suddenly becomes one of the next assignments, and everything starts to kick in gear. Luckily, our team was prepared for this to happen. We had good expertise from Marcus and Brendon in manufacturing. This was coupled Jacob’s and my eagerness to learn some of their skills. Our design was complex. This was bad because we occasionally got into situations that none of us were very certain about, such as the timing circuit, but once the extra time was put in, the reward was definitely there.

We also incorporated laser cutting, water-jetting, and 3D printing. These were all new ways of manufacturing parts that I had never used before. It definitely added to the learning experience of the class by learning how to use all of these devices and what type of materials can be used on each device. Once our bots were finished, the design expo was a great way to cap everything off. All the stress of finishing the machine was complete; all that was left was to see how our machine performed in comparison to the rest of the class. Fortunately, our bot came in 4^th, so the effort was definitely worth it.

There are two things that I would change about this class. First, I would recommend more CAD labs prior to the creation of the machines. We had a bit of lopsidedness when it came to experience on CAD, which slightly affected the difference in workloads between us. Secondly, I would recommend a different way to post assignments. There were many times when our group wasn’t clear as to when assignments were due and what exactly the assignments entailed. Other than that, I enjoyed ME 250 and definitely feel that I got a lot out of the class.

Thank you team Fluffy Bunny and ME 250 staff for making ME250 an informative and enjoyable class

-Jimmy Cofsky

Individual Reflection - Marcus Brown

This was probably my favorite class this semester. I enjoyed that this course allowed us to take all the knowledge and experience we’ve accumulated in our lives and finally apply it. The hands on aspect of the class was great and taught me a lot. I think the biggest thing this class teaches any of us is the distinct difference between designing something, and manufacturing it and bringing it to life. The main offender in this turned out to be not double checking each other’s measurements. There were several cases where we accepted or assumed a value, and when we got to manufacturing and assembly, came to realize we had to change several dimensions. Otherwise, we had things pretty sorted out when we finished the design phase. During manufacturing, it may have been the horrible lack of sleep, but the saying “anything that can go wrong, will” proved to be all too true for us. The few things that slipped past our attention and weren’t completely settled tended to blow up and cause more problems as we had to change things. This gave me further experience with handling problems on the fly, in less than ideal situations.

We designed a timing circuit to operate our roller and eliminate it as a controlled degree of freedom and allow it to spin during the match then shut off at the end. This circuit was an interesting experience, and has reinforced my decision not to be involved in the EECS world. It was cool, but I’m glad we finished it. Because some of the control modules apparently leak current to the motors despite no input from the controller, our tests of the circuit before and during seeding showed us it would trigger unprovoked. Because of this, we elected not to power the timing circuit during competition because we worried the roller would be more likely to cause our disqualification than it would help us. The roller was positioned to just barely touch the top of the balls, and it was powered through rubber bands on pulleys, so we were able to operate without the roller spinning. We did, however, decide to drive with the circuit running, after the final match, and were pleased to see that it operated as designed, though frustrated that we never used it in competition (a short clip of it is posted below).

Our team functioned very well together. We all got along, and were able to all contribute positively to the design and manufacturing of the project. We all stuck together and helped out during our disasters and worked to achieve our goal. We all remained focused and motivated through our endeavor and were pleased with how it turned out.

Something I would change for the future, would be having a central location for all assignment related notices, information, and links. It was very difficult to find information that I knew existed somewhere on the ctools site. It was especially difficult and time consuming to find links and important information that were essentially buried in the lecture slides.

Given the opportunity to do it again, I would probably elect a simpler design, as the detail in our design caused us to spend a lot more time working on it than a simpler design would have. Fewer small parts to machine and grind would have been a good choice. One major thing I would change to improve would definitely be getting more sleep. Also probably would have pushed the schedule further up to allow for more testing and diagnosing of problems. As a result, our defensive “decepticon” module was not as effective as we had planned, due to what we have agreed upon to be the result of minimal weight over the wheels, resulting in very little friction, and far too much resistance resulting from the length of the wire to its motors robbing them of power.

Thank you to my teammates on team Fluffy Bunny, for making this a solid experience, and thank you to the ME 250 staff for making it a fun, informative class.

-Marcus Brown

Final Bill of Materials

https://docs.google.com/spreadsheet/ccc?key=0AoKo-lBMy5cNdEpQNUdzRzVDVEhnSmlWbjRyX3diLWc&hl=en_US&pli=1#gid=0

Reflections by Brendon Sassmannshaus

ME 250 has been hands down the best class I’ve taken so far. It is also by far the most time consuming.

I am extremely happy I wound with a compatible group. My Engin100 groups were a mess. This improvement could be from more cautious selection, or that ME250 is comprised only of mechanical engineers. As a group, I feel we struggled with integrating less familiar team members into processes when one of us had paramount proficiency. I feel we did well organizing meetings, discussing what needed to get done, and conferring on our successes or problems. We could have chosen a less ambitious design to make the project a lesser degree of time sink, but that’s not “how we roll” on team Fluffy Bunny.

In the design process Marcus Brown kept extremely well broken down sub assemblies of all of our components to limit the complexity of each layer of our CAD models. I plan to keep this practice; it was a great boon. The laser cutter was an amazing convenience along with the Water-jet but not being able to press fit (esp. bearings) into products of either of these tools is limiting. 3-D printing was a large part of our manufacturing. It was cheap, versatile, reasonably fast, and took almost no supervision.

Our timing circuit for our roller was a really fun piece to design. We wound up not using it in the competition because we couldn’t trust the output of every control box being within tolerances (a small leaking current could trigger the latch and disqualify us) and the circuit went into very dark grey area in terms of the degree of freedom limitation after it timed out; the roller would get current whenever the trigger was picked up after the 70 seconds. This is pretty much signal encoding and it’s effects could be produced mechanically with two ratcheting mechanisms but upon asking Mark it seemed to be against the rules.

Similarly vague was the rule on ‘sand bagging.’ I heard during the competition that it had been ruled that you could sand bag for traction. The only other things I can think to sand bag for are momentum for a collision, and deliberately slowing your robot both of which seem pretty improbable given the speed of the competition. These rules should probably see errata.

I really enjoyed the course and got a lot out of it. However, I would recommend a better system for displaying assignments. Both on the spreadsheet and on c-tools the assignments seemed hidden in lectures and virtually non-accessible to sleep deprived students. This could be due to multiple instructors collaborating in the class.

To improve my performance in the course I might have chosen a less-ambitious project. Decepticon was a concise simple model in my mind so we went with it, but it limited our main bot to 2 degrees of freedom, had a lot of flaws, and in the end was a minor flop, though it probably could have been remedied by sand bagging the wheels and decreasing the length of wire or changing the gauge to decrease resistance.

Thank you to my team who stayed together through the 4 AM traumas.

A BIG thank you to the ME 250 staff who gave us such a great environment to learn (and sometimes make mistakes.)

~Brendon Sassmannshaus

Competition

We finished the design expo competition in 4th place and also received an award for best manufacturing.

We never ran the roller due to unprovoked triggering of the timing system during testing, and fear it may malfunction causing us to be disqualified. However, we ran it after the competition, and were pleased to see it ran just as expected.

Intro Video!

Here's our intro video for the Tech fair on Thursday:



The video was done with some some interesting rigs, kind of like this:

Top of the box with batteries as rollers, a wooden plank for the slider, and doorstops to change the angle of the camera.

Project Update

We have passed our MS9 due date and were able to run both Autobot and Decepticon on Monday (12/5/11). We have resolved the last few problems we have encountered which include the coupling between the motor and the drive shaft as well as the timing circuit. Decepticon’s sizing wasn’t exactly as expected, and since we had extra acrylic, we decided to laser cut a new body for decepticon to ensure that its defensive purposes will succeed. Today (12/6/11) we have been in the lab using the controller to test our machine’s maneuvering and review our strategy for the competition tomorrow. While testing Autobot, we learned that it was much quicker than expected. Therefore, we have decided after calculations and consideration that it would be advantageous and safer if for the seeding round, we go after all of the front balls (on both our side and opponents side) and then the balls at the base of the tower. With our speed, we should be able to pick up a significant amount of balls this way. Also, because we will be running unopposed, we will be able to get every ball on the front of the arena. This altered strategy provides a safer way to ensure that we get seeded high for the tournament on Thursday.

Our robot introduction video is in the works and will be ready to be played during the competition tomorrow. The final bill of materials is being produced and corrected, and everyone on our team is preparing for the tournament on Thursday.

Manufacturing update

We were not happy with the way the acrylic side plates turned out so we are now  remaking them. There was too much friction between the plates and the wheel axles so we have redesigned the plates to account for the problem.

A problem with the laser cutter caused us to ruin our acrylic plate so we traded for another and are currently in the process of cutting new side plates. We are also printing out Decepticon's body simultaneously. Upon completion of this print we will have Decepticon's pieces done.

Autobot's aluminum side plates also need to be adjusted slightly to better place the roller at the correct height.

Late night in the Lab

Printing a tensioner pulley. Had some trouble with the MakerBot, but it was fixed by Brendon and is now chugging along making more parts.




Also, testing the roller circuit. We've fortunately had the help of a brilliant friend of ours, Nick Clayton of the EECS type, who has been invaluable in aiding in the solving of our timer problems.

Top, sides and back done

The finished Top, Back, and the aluminum plate and acrylic Sides.

Waterjet finished

Our waterjet parts were finished this morning, so all that remains for autobot is to mill the bearing holes and finish grinding the right angle brackets we made using the kit 3/4 inch square stock before we can do our final assembly of the autobot module

Roller

Last night, we finished casting the roller and removed it from the mold. We have also finished laser cutting for our "autobot" module and only have waterjetting and milking the aluminum plate sides to accomplish for the module.

Body Construction update

Despite some problems with the laser cutter and its offset we are now back on schedule with the body of Autobot. The top was manufactured fine, but sometime in between the cutting of the top and sides the offset was altered, messing up the cuts and preventing them from fitting perfectly with the top and from fitting the shafts for the wheels. We fortunately had enough acrylic left over to redo the side cuts which are now as perfect as we can get them. We'll post pictures soon.

On a side note we've finished the soldering of the wires to the drive motors of Autobot and will be finishing those connecting Decepticon to Autobot tonight.

progress

tension pulleys and hubs (now assembled)


main pulleys/wheels and axles

casting the roller

Updated Schedule and To-Do list

We have updated our calendar with expected times of completion for the remaining parts that need to be manufactured and assembled. We are still waiting on our piece of 1/8th" aluminum from McMaster Carr. Once this is received, we will be able to complete the manufacturing and assembly of the body of Autobot. Because we were slow to receive this piece, our completion of the two machines were pushed back to the end of the week instead of our goal completion time which would have been mid-week.

Our to-do list includes (Overall):

- Finish manufacturing and assembly of roller (should be completed by tonight 11/28)

- Order last electrical parts needed for roller timer (11/28)

- Review dimensions of Decepticon, manufacture and assemble (11/30)

- Timer circuit assembled for roller (12/2)

- Attach motors to bots and attach lead wires to motors, everything electrical complete (12/2)

- Autobot and Decepticon completed (12/2)

- Blog post that includes recap of completion of both bots as well as video (12/6)

- Bots are inspected and held until competition (12/7)

- Debugging and practice maneuvering the machines (12/2-12/7)

- Individual Reflections (12/12)

*dates in parentheses are expected completion dates

First pulley and insert mated

Who needs a press when you have a text book and a table? It pressed in securely and centered. All good so far.

Team Logo

Most Critical Module Completion

We chose our drive train to be our Most Critical Module because without motion, our bot would not be able to carry out our strategy or get any points, which would lead to instant defeat. For our drive train, we decided to have a tread system. This system would help with traction when trying to climb the back of the arena, while still ensuring maneuverability. We were able to complete our engineering for our Most Critical Module by the due date of November 11. This included the solid models, part drawings, bill of materials, and analysis. From November 11 until November 21, more calculations were done, some specs for the parts of the MCM were adjusted, and we were able to manufacture 3 parts in time for the MS7 due date of November 21. The parts that we manufactured to our specifications and drawings were two sets of pulleys, which we made using 3D printing, and our tensioner shafts, which we used the lathe and drilling processes to manufacture. These parts were checked off in lab and met the specifications of the edited drawings.



As for future machining and progress on our bot, since we have completed our manufacturing for our MCM (drive train), the next component of Autobot that we are going to be working on is the roller. The roller will be molded out of silicon. The mold that we will be using for this is going to be made by a 3D printer to the specifications we desire. We will also be creating the roller insert, which stiffens the roller and transmits the driving force.






Here are two images of the components of the roller that we will be manufacturing.

There will be some difficulty due to Thanksgiving break, but our group has been good about communicating and determining times to meet that would cause minimal conflicts. Pictures of the molding process and construction of the roller component of Autobot will be posted shortly after the roller has been completed.

Calendar

Strategy

Our strategy is a complex one. We have two individual robots instead of a single one. They operate independently of each other to accomplish different tasks quicker than a single bot designed for both tasks could do them. The robots are codenamed Autobot and Decepticon.



Autobot:
The main bot. This robot is responsible for collecting balls and depositing them in the scoring hole; a task most of the bots are used for.It starts by picking up the balls on the front of the table and depositing them into the scoring hole. It then drives up the stairs to reach the balls on the upper portion of the table. After acquiring those balls, Autobot drives back down the stairs and deposits them into the scoring hole.
Autobot utilizes a drop-six belt drive-train, enabling the bot to drive with a large amount of grip (thanks to the belt) yet not scrub when turning. Also, the belt drive allows Autobot to climb the stairs with much greater grip and therefore speed by evening out the stairs and allowing Autobot to climb them like they were a ramp.
Autobot has a roller mounted on the front. The roller is large and easily compressed in order to mold itself around the balls as it sucks them in and grip them more securely. Autobot's underside is open, so that when it drives over the scoring hole it automatically deposits the balls into the hole.



Deception:
Decepticon is (dare I say it) deceptively simple. Decepticon is simply a plate a little larger than the scoring hole with the double gearbox mounted underneath along with a couple sliders in the front for driving. Decepticon is designed to drive into the opponent's scoring hole and block it upon Autobot's scoring of its first ball. Once Decepticon is in place, it can not be easily removed due to the sliders and gearbox preventing horizontal movement. Once entrenched, Decepticon is very low to the ground and most bots would simply drive over it, preventing other bots from simply knocking it out of the hole.
Decepticon is connected to Autobot through a line trailing from its rear end made up of four electrical wires to transmit power and control from Autobot to Decepticon. This wire will attach to Autobot with a spool, but will largely start off uncoiled on the table in order to allow for quick and easy separation between the two bots.