Winter Break

MIT's Independent Activities Period is good and bad.  Its good because there are a lot of opportunities and it gives students a break from regular semester classes.  Its bad because it lasts until February so that I won't graduate until June!  Regardless, I have been making the most of my time recently by keeping busy.  I was lucky enough to spend over a week at home which my family, which included Christmas and New Years.  Despite the festivities I somehow got some work done!

First off, I had to help my dad fix the tire on our minivan.  He teaches a high school auto shop class, so its really cool to see where he works and learn from him.  The front-right tire pressure was dropping to around 15 psi just days after filling it up, so we knew something was wrong.  After a quick inspection it turned out the culprit was a huge nail the van must had run over recently.  As I learned, the way to plug the hole is apparently to coat it in industrial rubber cement and jam a sticky, jerky-like stick in the hole.  My dad told me he's been doing it that way for decades, so it must work pretty well.  Once you let it set for a minutes, you just pump up the tire again, trim the ends off, and voila!

The ends wear off as you drive. Who said car maintenance was pretty?

I also finished up the CAD for my first electric vehicle.  Its not terribly exciting, but the project is suitable for a beginner.  I was inspired by the idea of wanting to make a scooter, but not having the materials or money to make an expensive one.  I present to you.... 2x4 Scooter!

Composed almost entirely of wood 2x4s, it has two 8" wheels, rear wheel drive via a brushless outrunner motor coupled with a timing belt, and I plan on making a lithium-ion battery pack to give it plenty of power.  I'll post the build once its finished!

Alpha Prototype: Cobolt

     That's right, the electronic bike lock I've been working on is done (and underwent another name change)!  Here's a picture of the prototype while opened, with the keyfob in the mechanical backup.

We named it Cobolt to signify strength, technology, and reference its locking ability with "bolt."  Mechanically, the team implemented the changes we needed to make it lock more reliably and securely.  Electronically, I figured out the RF transmitter and receiver bugs and helped program the microcontroller.  In the end, we had a working alpha prototype!  Here's our presentation given in MIT's own Kresge Auditorium, followed by the audience Q&A which features yours truly!

Presentation

Q&A

 

I really like the concept, and I don't think it would take that much more work have it be useable on a daily basis.  Who knows, you might see something like it on a bike rack sooner than you think!

SigmaSecure (Happy Thanksgiving!)

That's right, we had a name change.  Our 2.009 electronic bike lock has made a lot of progress since I last posted.  It's sleeker, more secure, and the electronics are a lot more developed.  Plus, it's no longer just a CAD model! After several weeks of hard work, we got an actual working prototype.  Here are they key features:

• One piece
• Spring-loaded hinge design
• Key-less entry
• Wireless communication via XBee
• Solenoid powered dead-bolt
• Button activated
• Backup key for mechanical override

The pictures from our technical review were taken down, but here's a Solidworks screenshot:

 

 I was in charge of electronics, so once we ditched RFID I decided to go with Xbee.  They turned out to be a very convenient solution with longer range capability.  By having an XBee on a low-power mode in a keyfob paired with an XBee in the lock, SigmaSecure can tell when you're nearby and unlock when you press the button. Even better, you can program XBee to signal when it is communicating with another XBee.  With this, we were able to trigger a transistor to allow our 9V battery to power the solenoid.  Since the XBee runs off about 3V, we simply used a voltage regulator to allow it to run off the 9V as well.

Next, we are looking into either using a latching solenoid or motor powered locking system to prevent some failures related to our spring-return solenoid.  Since the final presentation is only 3 weeks away, we are going to quickly make these last-minute changes so that we can end up with a polished product.

Now, back to the festivities at my girlfriends house...

SwiftSecure

With all of my class midterms behind me, I have recently been spending an excessive amount of time on 2.009: Product Engineering Processes.  Last week I helped present the 'assembly model' of my team's project.  My team is working on an automatic, electronic bike lock idea that I proposed several weeks ago.  To unlock your bike, you simply press a green button on lock, and an RFID reader senses if you're nearby.  If you are close by, which it can tell from a small wristband or keyfob device, a solenoid retracts a deadbolt and the lock swings open.  My job was to create a housing for all of the components and securely integrate them into one piece.  I did some renderings of the solid model we made, so here's one with the side cut away you can see what it would look like, along with the product contract which contains preliminary specifications for the product.

Sound cool?  I sure hope so!  Still have some concerns?  That's okay, because we still have a lot of work to do before we're finished, even though the class deadlines are approaching quickly.  Until next time!

DragonBot Finished!

Between class, homework, applying for jobs, and other various duties, I finally found time to finish DragonBot!  Here's how it works:

DragonBot has two continuous rotation servos, one on each side; this allows it to drive forwards, backwards, and turn on a dime.  A large, high torque servo found in the middle of the base is used to spin a rod that functions as a pulley.  This pulley has a string tied to it that is hooked to one of the bars in the four-bar mechanism.  The four-bar mechanism consists of four bent strips of steel, and is supported by two 16" lengths of aluminum box extrusion.  By using the pulley to draw in or let out string, the grabber on the end of the four-bar mechanism can be raised and lowered.  The four-bar mechanism allows for the grabber to remain parallel to the ground as it goes up and down.  The grabber operates using a continuous rotation servo with a gear on the end.  The servo is on a linear bearing so that as the gear spins, the whole thing moves along a rack, allowing it to pinch objects between two platforms.  If that was a bit confusing, just take a look at the photos below.

I had to make a few modifications from the original design when it came to actually building DragonBot.  Firstly I only had smaller Parallax servos instead of the stronger SpringRC variety that I wanted to use for the driving wheels and rack and pinion gear.  The press fit I had designed for holding the grabber onto the four-bar mechanism was about a millimeter too big, so I layered some tape to tighten the gap.

In terms of things still left to to, the most important one is to machine grooves for e-clips on each rod so that everything doesn't slide around.  I could also come up with a more permanent and secure fix for mounting the grabber on the four-bar mechanism.  I also would like to add rubber pads on the rack and pinion grabber to increase its grip, allowing it to pick up heavier and more complex shapes.

All in all, I think DragonBot is a great success.  The grabber works perfectly and the the lifting arm is extremely strong.  Even so, if I attached the string on the pulley further back on the four-bar it would increase the mechanical advantage by about 3x!  Check back soon, I want to get a video of DragonBot in operation!

Back at the MIT + DragonBot Update

As the title mentions, I'm back on campus!  I've been very busy with my fraternity's rush, so not much work has gotten done, but I have at least made it through my first day of class!  Hopefully I'll have time to go into Pappelardo to make the rest of DragonBot before the end of the week.  Since the grabber is done I just need to make a four bar mechanism and a base with wheels and aluminum box extrusion uprights.  Oh, and I also need to figure out how to lift the massive grabber.  To cut back on weight the four-bar will be made out of steel sheet metal which I'll bend to make more rigid.  I'm hoping this will be strong and very lightweight.  The base will be made out of ABS and will have standard 2.007 continuous rotation servos and wheels for front wheel drive, with a caster wheel or two in the back.  To stabilize the tall aluminum box extrusions I plan on bolting on aluminum brackets to attach it to the base.  Here's what it will probably look like:

At this point I decided to do some calculations in order to get an idea of how much force I'd need to lift the giant arm.  To get the mass of each part I applied material properties to each model in Solidworks.  The four-bar is 16 inches long (0.41 meters) and should weight about 0.94 pounds (0.43 kg) and the dragon wing pieces weight about 0.12 pounds (0.05 kg) together.  The grabber should weigh about 0.77 pounds (0.35 kg) and is another 2.5 (0.06 m) inches out.  Simplified, this is 0.34kg at 0.41m and 0.35kg at 0.47m.  This means that all together I'll need to apply at least 0.3 kg*m of torque at the top of the box extrusions.  I was planning to use a Vigor VS-11 servo which applies about 0.16 kg*m of torque at 5V, so I'll need to at least get a mechanical advantage of 2x; 4x would probably be ideal.  If I use a 1/4" diameter pulley attached directly to the VS-11, I'll be able to pull with a force of 25.2N.  To get 4*0.16= 0.64 kg*m of torque I'll need to extend the four bar mechanism to stick out 1 inch the other way and attach it to the pulley with a string.

Well, that's about it for now (I am le tired).  And don't think I forgot about my other not-so-secret project, I promise I will reveal the details in my next post!

End of Summer + DragonBot Intro

It is the end of summer, and school will be starting in less than two weeks. I recently finished working on Robocon, which was a great experience. For details on that, check out the 2.007 and Robocon link, also on the left.

Before Robocon ended I was working on 2.007 style robot I have dubbed DragonBot. It was basically a more legit remake of a demo robot I made as a 2.007 UA in the spring to pick up a toy police car for the contest. I already waterjet some parts out of ABS plastic and assembled the grabber, but it is in my room at school. I am currently at my girlfriends house so I'm not able to work on it, but I should be able to finish the robot shortly after classes start. However, I still have my cad files! Here is one of the grabber, which is a rack and pinion.


The design allows for a strong grip due to the small radius of the pinion gear, applies force only in the same direction regardless of object size, and can let go cleanly without applying any torque to the object its lifting. This is key for precise placement of objects, which was critical in the 2.007 contest. Most of it is assembled using t-nuts (as seen in the lower left corner of the picture above), which I learned about from Charles. They are really handy when waterjetting parts; all you need to do use bolts to piece things together (I can explain in more detail if anyone is interested). So why is it called DragonBot?
That's right: waterjet dragon wings on a four-bar mechanism. The grabber simply slides into the slot you see on the bottom.

Well I guess that's all for now. I am also starting to work on a new super-secret project which I will reveal in my next post (lies), so check back for that later (or subscribe)!