PF-Arduino-WeDo Hub (Found on Cuusoo)

[Junpei]

Today I found this on Cuusoo: http://lego.cuusoo.com/ideas/view/49046

How awesome is that? It would open up so many possibilities, the first one that comes to my head is a differential gearbox!

Maybe we can persuade TLG to just make this instead of waiting years for it to <possibly> get 10000 supporters, it's the best idea I've ever seen on Cuusoo.

[DrJB]

Sounds neat, then you might as well do a Raspberry-PF box as the latter is supposedly 'non-profit'. Yet, slowly that idea is migrating towards the Mindstorms product line. How do you ultimately draw the line and differentiate between the two? Somehow TLG has kept 2 separate product lines: 1. The Remote-based PF (human-controlled), and 2. Mindstorms Robotics, that where the toy interacts with its surroundings through a set of 'smart' sensors. They tried something in between, as in SpyBotics or StarWars Scouts, but with no lasting success (in my opinion at least). Lately I have seen many 'mechatronic' Cuusoos ... and while those are neat 'ideas', I'm afraid they'll remain just that. Do not get me wrong though, I would be pleasantly surprised if those come to life.

Edited October 16, 2013 by DrJB

[CP5670]

Somehow TLG has kept 2 separate product lines

I never understood why they did this. In the RCX days, Mindstorms and Technic were based off the same pieces and were essentially compatible with each other. Today, Mindstorms uses different connectors, motors and even color schemes from PF. The WeDo system brings some of the Mindstorms functionality to PF, but is highly overpriced and has its limitations.

[Junpei]

I agree with CP5670, I liked that the RCX and 9V were compatible in every way. This would just break the barrier between both, not to mention add many new possibilities.

[JopieK]

A very nice idea!!!

@DrJB: Raspberry is a totally different system (much complexer to program). I teach robotics / embedded systems in high school and college so have some insight knowledge on different microcontrollers etc.

One might consider a Teensy 3 as backbone however (prjc.com): it is a very powerful ARM processor and compatible with the Arduino environment and language.

LEGO will never do this I think since it is way too open, makes them less profit... I'm currently working on an Arduino LEGO train with Xbee, if we (my students and I) get it working, I hope to post more details of course... the programming part is the largest challenge if the hardware is right (new circuit board is hopefully arriving soon from China ;)).

B.t.w. WeDo is very nice, but indeed expensive, the concept works great for primary education (especially the younger kids) though! It just works...

[DrJB]

@DrJB: Raspberry is a totally different system (much complexer to program).

Good point, though I never implied it was easier ... but I emphasized it was 'cheaper'. I never cease to be delighted as to the 3rd party OS/apps you can find for various hardware ... be it LeJos or else.

LEGO will never do this I think since it is way too open, makes them less profit...

I'm afraid I have to agree with you on this ... Edited October 15, 2013 by DrJB

[JopieK]

Raspberry PI is more a small computer platform where Arduino is a programming concept + a hardware concept. Disadvantages of the normal arduino's are that they are not very powerful, the Teensy 3 by Paul Stoffregen is programmable using Arduino code but also very powerful (the uC is 48Mhz and will even be even able to run at 96Mhz, a lot of PWM pins, three uarts, etc.), without the need for an operating system. That is one of the disadvantages of the EV3 of course: boot up time is very slow because Linux needs to start first.

[DrJB]

Raspberry PI is more a small computer platform where Arduino is a programming concept + a hardware concept. TheTeensy is programmable using Arduino code but also very powerful ...

... You got me curious now, I've had a raspberry for a few months now and there are many add-on IO boards available. I have 'seen' some projects on-line where people are using the Rasp to control various mindstorms transducers (ipo NXT/EV3). I just looked up Teensy ... seems like a new 'player' on the block, will need to look further and explore potential. Thanks.

[msx80]

Awesome project, i'd love to program with arduino..

Also, the separation between PF and Mindstorm has always bugged me. There shoul be one single line with different capabilities. Now i have 6 motors but i can only use 3 in a project, either PF or EV3. Sucks.

Also, i like that it's small. EV3 board is HUGE (and the components too).

[pmichiel]

Mindstorms & Raspberry PI = BrickPi: http://www.dexterindustries.com/BrickPi/

The way I see it; PF is for remote controlled models. For this you need strong motors with speed. You need connectors that can just pass voltoage and a lot of current.

Mindstorms is for Robotics. For this you need sensors and motors with precision control, i.e. the interactive servo ones. For this you need connectors that pass current but are also able to transfer sensor data at the same time.

To me it doesn't seem like a marketing trick but a sensible decision based on the completely different product ranges.

[Jim]

This is a nice idea!

I was hoping for Mindstorms to blend more with PF, but there will probably be a very good reason TLG decided not to do this.

(maybe pmichiel is right and the answer is really simple)

A separate power supply, with a small controller, possibly one without a screen, with the option to connect a screen. This way it would all be way more modular. The controller should have built in BT and WiFi and should come with USB host and client port.

Overall I am very happy with both products, but this would have been great!

[Rob Klingberg]

Very interesting idea! Although the NXT today can communicate via I2C natively, which Arduino can also use, so why not make a NXT-to-Arduino I2C connector? Then people could use whatever Arduino board/flavor they wanted.

If you have manufacturing/design/electronics experience and really want to bring your idea to life, my suggestion would be to not wait for Cuusoo but instead go directly to Kickstarter.

[rambambulli]

Hi guys,

Nice to read your comments on my Cuusoo Idea. Lego deleted it from the Cuusoo site because it didn't match the new Cuusoo Guidelines.

I started a blog investigating if I could make my own Lego Power Functions Hub. It is a fun project so far.

I have to add a youtube movie to my blog to show how it works. I'll add it soon.

http://impyourlego.b...lego-power.html

and

http://impyourlego.blogspot.nl/2013/12/project-1-lego-ir-transmitter-with.html

Please check it out and let me know what you think of it!

Thanks,

Jeroen

Edited December 22, 2013 by rambambulli

[Moz]

I was hoping you'd go the other way, and use BLE to talk to an Arduino in the Lego-compatible controller brick. But this is a much easier first step, especially for someone with no electronics experience. I have a pile of RFduino stuff here that could form the basis of a compatible kit, I should play with that. Hmm.

[rambambulli]

Thanks Moz for your reply!

I actually started with a BLE device. I used a TinyDuino + BLE shield. It works.

So I didn't switch because to the Imp because the BLE solution is more difficult.

The thing I really like about the Imp setup is the simplicity. I switched because the Imp is easier. If it is easier for me it will be easier for most people. The hub should be understandable and affordable (Imp costs only $29) for as many people as possible.

The main advantage I think is you can use any device with an internet connection and control your model instantly. The user interface can be extremely easy (really basic html, even only a url will do) or more complex (processing.js, etc.). You can use iPad's or iPhone's Gyros at the same time you use the keyboard of your PC. There are tons of usable codes out there.

Another fun thing is you can control for instance 4 Lego Crawlers with 4 friends using only one hub. Or control a dirt truck and a crane with 1 hub. You control the truck with your iPad and have your PC calculate some recorded movements for the crane. Endless possibilities.

I'll try to show it on a Youtube movie.

I realize now that there is one major thing I have to add to the CONS. You always need a internet connection. Outdoor situations are becomming more complex. I managed to do this by creating a wifi hotspot on my phone.

Moz thanks again for your comment! It helps me. Maybe I have to take a look at my TinyDuino again

.

Jeroen

Edited December 22, 2013 by rambambulli

[Rob Klingberg]

I looked at the Imp as well for a wireless platform I am developing, but decided against it for the very reason you mention: you always need an Internet connection. While we are able to count on these things more and more frequently in more and more locations, they will not always be reliable, especially in places like convention centers where LEGO shows are held. In these situations, even Wifi cellular modems may not be reliable. Plus, isn't there some sort of recurring monthly charge to use the Imp's cloud-based service? I may be mistaken here, but if there was a charge, it would seem to add to cost over time. I've decided the better way for my application is to use BLE as a proxy to cell phone, which wouldn't need an Internet or cell connection to allow interactive programming and control. Plus the cost for BLE continues coming down, and more and more phones have BLE capability. Plus it is great for power-- anything that relies on Wifi (even the Imps) will consume many orders of magnitude more power than BLE.

Moz, I am curious to learn more about how you will be using your RFDuinos...

[Moz]

I wasn't going to use them with Lego at all, I bought them for the BLE as a sensor platform. I have a project where I'd like a network of environmental sensors, powered by little solar panels and ideally not wired. So the one-chip BLE setup in the RFduino seemed like a good easy way to start playing with mesh networks. But there is at least one board that will deal with large currents, so I could see how it copes with an XL motor. I'm actually thinking of doing a basic feedback setup using an analogue input and switching the output to get PWM, but I'll have to see whether it works at all first :)

For me, the goal is something compact and RF based, because IR is annoying with larger models and at exhibitions. I'd also like better control than the Lego IR gives, and ideally the ability to run off a 4S lithium pack so I can always supply 9V to the motors. That doesn't seem impossible to me, and with an Arduino you've got so much processing power available that that part of it is a bit of a joke. You could run a FSM looking at 4 outputs and 4 inputs at 100kHz or more, on a few mW. Driving the output transistors would actually be your major power cost.

But not this week, I'm kinda busy.

[rambambulli]

hmm You guys make me really doubt about my Imp project.

[Moz]

rambambulli, your project is a good start, don't get me wrong. Especially now that you have the parts, it's worth completing it. And you will get the phone-control part, which is what many people want. You could even build it into many models and phone-control it.

My personal issue here is that I don't want to start down the "looks like Lego" path if I can avoid it. In many ways the whole point is to work within the limitations of the medium, I've build computer-controlled machines and radio controlled toys before, but both of those are really, really expensive hobbies. Lego is cheap by comparison, but only if I stick to Lego. As soon as I start saying "add some electronics and..." the money starts evaporating. And getting a compact circuit board with big enough driver transistors to handle, say, 4 XL motors, plus a battery, built into a Lego-compatible box is not going to be cheap. I can get the proto 3D printed, and DIY the first circuit, but then.... what? Suddenly I'm spending lots of time, and lots of money, manufacturing hardware. Again. My partner would kill me. Again :)

Edited December 23, 2013 by Moz