After working with FEZ now for a bit, i have decided that it really will be a great solution for some in house projects - starting with test fixtures. I work for a really big company that makes big hardware. My plan is to put together a class of 5 two hour sessions with lecture and lab and give each student a kit with a FEZ Panda and a set of accessories. [Management just gave approval in the middle of typing this]
The students will be seasoned electrical and mechanical engineers - people familiar with motion control, environmental measurement, vibration and sound studies, power consumption experts, etc. Most of them will not be software engineers, but some will be deep embedded software people. i am planning on 10 students per class (at least at first). I may have as many as 40 run through the class before I am done.
The goal is to give these engineers a way to do their research and refinement inexpensively with off the shelf components - and some custom components when necessary. This would replace some of the custom PCB design and LabView stuff we do all the time.
I would really like to hear any input or advice - especially if you have done anything similar.
Well it sounds like all of these guys are experienced in the field, but if they are not(you would have a better idea), I wouldn’t over look the importance of laying out the programming side properly. Take a bit of the time, if not in handouts to explain “Classes” and “Methods”, (private, public, etc), naming objects and such.
For someone that has been in the field of making things(manufacturing/machining) with a bit of automation, computer programming is Greek to me, even though I can program a CNC machine like no one else, computer, Nope.
Mike, thanks for the input. I will be sure to focus enough attention to the programming aspect. I will post a course outline, kit specifics, and all that stuff on Fezzer when I get it ready. I expect the first class to run in early December, but that may depend on availability of parts for the kits.
I have been working on something similar for a client of mine Bob, however i can’t say anything about it publicly. If they are not used to working with micro electronics, i would suggest teaching doing perhaps 20 minutes on programming, then 20 minutes on using the programming skills they’ve just acquired to work with some device.
For instance, on your 2hr tutorial:
Part 1 of tutorial: Cover basic syntax, control structures, math and variables.
part 2 of tutorial: Connect a temperature sensor, convert the voltage to degrees, then blink the led if the temperature goes more than 5 degrees above room temp. You can get the sensor to go above room temp by pinching it.
Part 3 of tutorial: Explain how analogue and digital inputs/outputs work now that they’ve seen an example of it.
Part 4: Explain collection types, looping, more maths
Part 5: Build on the temperature sensor to add averaging, perhaps the past 5 minutes history, etc
Basically, give them a small bit of something to learn, apply it straight away before they forget it, reinforce it, build upon it.
Mark, that is along the lines of what I was planning (it is actually 5 sessions each two hours long), but I had not considered the advantages of using a temperature sensor as a building block. That is a great idea. I will be covering digital I/O, displays, analog in and out, motor control, relays, etc.
All of the labs will use exercises related to our real-world application. These ideas are great!
Thanks.
I just chose temperature sensors because they are ultra cheap and ultra simple. Not much can go wrong with a temperature sensor. Rooms will be cooler than peoples temperature, so it’s a nice easy way to see temperature changes immediately (ie: hold the sensor to warm it). Most other sensors tend to be fairly subjective.
Just be sure that if you show anything new in code (look at all your code and see if there is any statement, structure or way of doing something you haven’t shown in a previous section) that you explain it… why you use it, when you use it, how you use it, what it does and why it does it.
If you’re showing control structures and you have an intelligent audience (who are not going to get too confused) consider showing them alternate ways to use those structures. ie: make a for loop act like a while loop, make a for loop increment by 3 or bounce back and forth rather than just incrementing.
Tell people that the var keyword was implemented by the devil, as were underscores in variables names. Declaring a variable with the var keyword and an underscore in the variable name is likely to make space time collapse. Especially if you have a cast on that assignment (constant issue here at work… lazy buggers) eg:
var _My_Private_VAR = someOtherVar as string;
it’s always a good idea to take something that people are familiar with and use it to teach new things. If you have sensors or devices you already use at work, use them for your examples once the coding skills get to the point they can.
OK, I am making progress - management gave approval. Their only concern is how we will run enough sessions to get everyone trained who wants to. Everyone seems to see that we will save on our R&D budget next year by building rapid development test fixtures the way I am proposing. One department came by already to talk about a cable test application that they want to build next week.
We placed an order for FEZ Tinkerer kits for the first class this morning. I have managers fighting over slots in the class. I am getting the first class filled up today. I just hope all of my kits come soon enough to get my teaching staff trained.
The reason for using the panda is because this first application is for one-off test jigs - which we build a LOT of. What they need is a simple way to do some digital and analog I/O and turn some motors and relays. Panda should handle that fine.for training and first applications. After they start getting used to the idea i will move them up to something more capable. For now, I want them to get the building block concept down and the Arduino compatible form factor should work great. I want them to start thinking smaller - but more numerous.
Hi Bob
I probably be holding a lecture at the University of Oslo next year and also had plans for using the Panda tinkerer kit. As it turns out there is a lot of soldering to do before you have the kit up and running. So we desired to go with a regular Panda and some breadboard stuff instead.
The people I will be speaking to are biologist and not that familiar with electronics and programing (thank God) but hopefully they learn the basics.
Geir, I considered the soldering aspect, but fortunately I have a team of technicians who can get the kits built up for me before the class. The thing that sold me was the preorder price - I get the panda, the board with SD card socket, RS-232 Connector, and the enclose for $10 more than the panda alone. It may be that the next batch of students may use a standard panda and a different enclosure - we will see. I will be very interested to hear what you present.
We have realized that some mid-level-users would love to use this kit but do not want to do much soldering so this was pushed back to our engineers to see how they can make this easier to do.
I have 2 classes filled so far - about 25 engineers. The courses filled much faster than I expected and i had to add a second class. I just hope all of my hardware arrives soon 8)
but hopefully they learn the basics.