Days 22 to the end

The long Memorial Day weekend was spent mostly in the woodshop getting our beautiful POD ready! David and Erica drilled some holes, chiseled some channels, and just got dusty! We had one minor setback in that our drivetrain failed (i.e. one of the wheel hubs started slipping on the wheel), but a day in the shop turning our hubs into double D shapes solved that! No more wheg slipping means great driving!

Ramanan and Rishabh continued to debug the code, and get it all ready for POD and ROAMER integration.

After the weekend, and 1 or 2 all nighters, we finally began to integrate. Integration was tough, and we blew multiple fuses ( forgot there was a metal screw underneath one of our boards....woops), but in the end we got it all working! Grading day was a bit hairy, as we were scheduled to go last and working right up until the last minute, because our ROAMER had some random disconnect issues that we were trying to debug. But we made it through grading! Ed really enjoyed our rocking horse :) He said we took the cake with entertaining PODS! #yes. #lifecomplete.

The public presentation also went awesomely, as the audience obviously loved our POD :) grown adults riding the rocking horse are super entertaining!

Another quarter of ME218 complete. The last quarter for most (all?) of us....unless Erica decides to take ME218D. It's the end of an era, but we've had a great (often sleepless, but still great) time. Thanks Ed, for the memories!

Days 21 through 23

It's finally getting there!

We got all of our inputs working! The accelerometers, force sensors and the buttons from the Pod all issue commands to the Roamer and it responds - brilliantly!

Going into the long weekend ahead, we need to get the stuff mounted on the Pod and get everything set for grading.

Days 19 and 20

Coding. Again.
So we discovered (much to our annoyance) that the code we had used for the previous checkoff was super buggy. Two days of troubleshooting later, the Roamer came around and responded nicely to the commands from our Pod as well that of others. We hadn't implemented all the controls on the Roamer yet, so we couldn't make a full fledged test yet.

The Pod, however, continued to be a pain. Since it had the C-32, we found it super hard to debug. We used a whole bunch of LEDs to figure out the states in which the code remained stuck. But whatever we did only seemed to make the problem worse. We're late for checkoff (again :-/), but hopefully we'll sort out the code by tomorrow.

David's been picking the actual Pod hardware apart. Its in pieces now. Literally.

Days 15 through 17

Circuit-making

So we decided that before we go into the next check-off, we needed to have all the hardware ready. Previous experience had taught us that we would save a whole lot of time if we started testing our code on soldered circuits in segments and not use breadboards as circuit prototypes.

Hence, we decided to invest quite a bit of time in designing the layout of our circuits - especially in determining the size of the board, the wire routing and in the placement of the molex connectors.

David continued to pummel through the SPI code to get it to work with the E128 while Rishabh checked the ability of the Roamer to run at different voltages and calibrated our servos.

Days 12 and 13

Coding.

Debugging can be a pain when you are coding in C on a C-32. And if the C-32 so decides to use those life-saving printf output lines as its Serial Communications Interface lines, you can consider yourself to be pretty much screwed.

Somehow, we got the communication going just a while back. Our code makes the Pod to issue a connect command to the Roamer, lighting up an LED in the process. When the connection is successful, another LED lights up. The Pod then sends back a command and the Roamer lights up an LED in response.

Hopefully, this is going to continue to work!

Day 11

Communication between the C32 and E128 has been a priority and the coding for it is in full swing. On one of the more hectic days of coding, the Pod finally arrives. David comes to lab with this

Pod - Ramanan calls it "Horsey" :P

The coding continued but at least things were a lot more relaxed.

Day 10

Getting the SPI up and running has been a priority apart from soldering circuits. The XBee boards are ready and so is a drivetrain board.

All the pin out diagrams have been finalized too.

Pinout Diagram

Motor Drive Circuit

XBee Boards

Day 8 and 9

The design of gripper was the obvious next step and it was decided to make a really simple gripper that has 1 degree of freedom and not just latches on to the grips provided but can also be used to lift the survival chamber. Thus a two teeth design that moves up was considered. However considering the accuracy in positioning needed for two teeth, a more robust design using comb was considered.

Using this, the bot would be gripping on to things, initial testing showed positive results.

Final design includes the drive-train from 218B project with Whegs, 2 servos to control camera position and the gripper and a sensor to sense the cage/survival chamber. Different options for sensing are still being evaluated.

After finalizing this, a CAD model for the roamer was developed and using laser cutting final roamer was made.

After laser cutting and gluing parts together the final roamer design was ready. Here are some pictures-

Whegs

Final Roamer

Special Mount with a Lazy Susan

Comb Gripper

 

 

 

 

 

 

 

 

 

Day 5 and 6

Looking at the uneven terrain that the bot had to traverse, different options were being considered - Tracks (not great for turning), 4 wheel drives (turning issues again) and big monster truck wheels. We suddenly remembered about one of the designs a team member had worked on before WHEGS !!

Whegs Design

That is what Whegs look like, they are basically Wheels + Legs inspired by insects and work really well for uneven terrains.

The team decided to test them first and modified a 218 Bot to use Whegs and test the bot out. A video will be posted shortly.

Platform used for testing Whegs

The Whegs worked perfectly on both surfaces and the interface between hem, the bot went forward, reverse and even turning in place. A three wheel design was preferred as the wedge contact ensures that both the powered wheels are always in contact with the ground.

Final roamer sketch

Day 3 - Update

Movement Software Description

Gripper Software Description

Day 3

We decide to use a PIC on the roamer to carry out both SPI (with the SPECIAL) and SCI (with the Zigbee). The PIC on the roamer will be assisted by a C32 on the Roamer while there will be an E128 on the Pod.

Roamer electrical components:

1. 4-5 motors

-2 for driving the roamer

-2 servos for the hook

-1 for tilting the camera (maybe)

2. SPI lines for communication between the PIC and the SPECIAL

3. xBee

4. 1 HC12C32

5. On/Off toggle switch

6. Switch to determine which Roamer it is (1/2/3)

7. LEDs to debug. Small protoboards for each circuit

8. Batteries

9. Voltage Regulator circuits

Pod electrical components:

1. Sensors

-Accelerometers - to detect rocking motion and convert to movement of the roamer

-Force sensors - to detect tension in the cable and carry out turning motion of the roamer

-Limit switches - on whip for executing the grab action

-Toggle switch - to change roamer direction

2. Switches

-bot select

-on/off

-camera control

3. Indicator LEDs

4. Batteries

5. Voltage regulator circuits

6. xBee

7. E128

8. Servo for showing the battery level

Day 1

Teammates: Myself+Rishabh Bhandari +Ramanan Sampath +David Zhai

5/1/13 - Day 1

Sitting in the dark and danky GCC, we all wonder at what to name our ME218C project. Ideas begin with what theme to model our POD after:

1. Erica’s bike...she’s already making one, so why not have audience members pedal to make the robot go, steer to turn it, and ring a bell to snatch?

2. Saint Bernard dog doll/large stuffed animal...pet the dog to snatch!

3. …..and somehow all this resulted in us deciding that making our audience members sit on a ROCKING HORSE to control the robot would be the cooliest.

Quotes of the night:

“What if we just make it poop everywhere??” - Ramanan referring to how to display a loss of energy. Ramanan has a strange fascination with poop. It’s disconcerting. We are worried about his mental health. Obviously his strenuous course schedule has already started taking its toll.

Rishabh is a debbie downer and not excited by our cool team name.

Ramanan's contribution for the "if BOT already controlled and another POD tries to gain control" scenario:

TEAM WHIPLASH!