I have a class of fifty students that are doing an mbed intro lab with several cookbook I/O setups this week and next. I was thinking of assigning a mini mbed design project after that. They work in teams of two for a couple weeks, so they should be spending around 12 or so hours on the project on average. Most are forth year computer engineering majors, so they know digital hardware and C. I like to let them each pick their own different projects, but often times they need other ideas and I typically give them a list of ideas to think about. We have an assortment of robots, hardware, breakouts, and sensors around along with a small budget. If there are some really interesting low cost devices available on short notice we could purchase a couple for projects. Hopefully some of them will wind up posting their projects.

Some of the ideas I have are:

Mbed control of an iCreate Robot and some other robot kits (arm, sumobot, biped)

Mbed sending SMS messages using a GSM modem with GPS location info

Mbed remote weather station using networking with a web page interface

Mbed with solar cell power - try to reduce power using power control library

Mbed reading RSS network news feeds on LCD display

Mbed Internet radio – with some extra mp3? audio hardware (any ideas on what?)

Mbed LCD or LED Clock Display – with automatic network time sync

Mbed using CMUCAM or another camera chip. Perhaps on a pan tilt servo setup.

Mbed Speech Synthesis with some speech IC (don’t know which one would be best)

I should have twenty or more teams and could use some more ideas. Does anyone have some novel and interesting ideas for student mbed design projects or a new low-cost IC or device to interface to add to the list for the students?

Whatever the project, it would be a good idea to incorporate some concepts they should be familliar with. For instance.. digital PID implementation (for an inverted pendulum for instance) or some digital filter implementation (i.e. signal conditioning in the digital domain). Or A distributed computing platform using mbeds..etc.

I guess what I am trying to say is to come up with projects that require some research and thought and that would be helpful in their careers (such as instrumentation, DSP, communication, control systems and other such real-time programming). For instance, the GPS over SMS could be usefull and a fun project, but I am sure you'll be able to find something on the net that does exactly that with the code already written. Students will often take the easy way out and copy someone elses work. I should know.. for our engineering design project, the majority of the groups ended up peicing together and porting code that has already been written by someone else and as a result, they learned only that.. porting code..

Sorry if I kind of went on a rant.. I've worked with some .. lazy.. students.

You are right, I left motor control with encoder feedback off of the list and will add it. I was checking to see if I had the parts around for it last week. 

Have you seen a low cost off-the-shelf way to do the inverted pendulum setup? I would need several setups given the numbers of students and a reliable source for spare parts to keep them working. I have seen some of the commercial setups for it for school labs such as http://www.quanser.com/english/downloads/products/Rotary/SRV02_ROTPEN_PIS_031108.pdf, but the price ($2000) is kind of a budget buster.

Students typically write code now by searching first on Google. I watched a group sit down and start a simple lab by searching first on Google, but all that came back was the URL to the class lab assignment I had just handed out to them.

They have a longer term semester long sr. design project later on.

I haven't really looked into actually obtaining one.. The ones I've seen were either some fancy kit or they were put together from seperate parts. And I am not sure diy is a good way to go if you need more than one. Oh, and idea that just popped in my head is a laser projector.. I believe the galvanometers are (relatively) cheap to get. See http://elm-chan.org/works/vlp/report_e.html for example. This design is mostly analog, but you can replace the PD controller with a digital control scheme.

This is one of the inverted pendulums they use in the embedded systems lab run by a friend of mine at ESIEE near Paris:

They have about 10 of them, designed and built in-house.

They are quite simple, and use a toothed belt driving a carriage on a couple of rails. The pendulum angle is measured by a (rather expensive) precision pot. The students have to model the system before they are allowed to work with the actual hardware. I could probably get a set of drawings and other details from my friend.

Hey!

Hey, the laser projector looks awesome, I want to make one!! plus, red diode lasers are pretty cheap nowadays,

and actually you/they could use a laser pointer, which is pretty cool and very cheap and

isnt going to blind anyone. with the laser and maybe an ldr with an appropiate filter, it could

be interesting to try to build an optic communication system ... based on i2c, spi or even morse patterns!

lol, go and search morse in google, you lazy student!

On my time, I hacked a Robosapien! that was lots of fun! ....

U can also hack an analog servo, opening it and wiring the signal from the potentiometer, that way

and having in mind it uses a proportional controller, measuring the time it takes from going from

one point to another, you should be able of, more or less, estimate the torque load in the servo... never

tried it, and is going to be tricky for the angular relationship but if they know enough control theory,

could be worthy. anyway I think for the students the most important thing is to enjoy learning,

otherwise is going to be useless for them.  If any of them is fond on cars they can have a go,

trying to implement the ignition mapping of a two stroke motorbike or a car or something like that,

so the can also learn, CAN/LIN, check this out, maybe you could find something similar for the

course http://www.motostudent.com/index_english.htm

Just remembered another cool project!, once I had to make a switch, with

a microphone, so when you whistle from grave to acute, you switched on a motor,

and when you whistle from acute to grave you switched it off, that way I just

expend lots of time working with time samplings, FFT, and memory allocation,

and actually was quite cool!

I would be interested on a estimate of what parts would cost for the French pendulum cost. I would probably need to add it to my budget wish list. I saw of video of a 2D one at a recent trade show by Beckhoff and used to show it in class, but the link is dead now.

I also found a paper about a remote video pendulum lab setup at a German school - you watch the live video of it moving after downloading your control setup. http://med.ee.nd.edu/MED11/pdf/papers/t7-129.pdf

That one is going to be a little longer term than a couple weeks.

Laser projector project looks nice - laser safety might be an issue given all of the other students likely to be in the lab. If you use legal laser pointer power wonder if it shows up?

It might as I had another project where students made a 3D scanner using a laser level (red vertical line), web cam, and servo. The vertical line is projected at an angle to the wall and anything it hits in front of the wall moves a section of it. They used the displacement of the red vertical line seen on the webcam to compute distance and scanned with the servo to develop a 3D model of the closest object.

http://users.ece.gatech.edu/~hamblen/489X/f05proj/laser_scanner/video.avi has a video of a scan

CAN bus looks interesting, I had some students try it a couple years ago, but they had a hard time finding all of the details about commands, codes, and OBCDII setup. They wanted to build their own OBCDII code scanner. A couple of them were using PC based CAN display tools for the school's motosports car.

The problem with the laser scanner is the position feedback. The galvos are not easy to make. I wonder if general servos are fast enough... Probably not eh?

I was at a conference in Argentina earlier this year where students got an mbed powered car with light sensors on the bottom.  Their task was to write firmware that would allow the car to follow a dark line printed on a sheet of paper.  At the end of the conference they had a race to see whose car was the fastest.