sephiroth117

For connecting a lego to a raspberry I currently use a buwizz 3.0 and a battery powered pi 3 A+. That's really a neat addition for Rpi and really in line with their STEM/education-first philosophy ! A better support for battery would have been neat, the only thing missing IMO, the big jack connector is not really ideal for that but it's possible with it I guess

Thank you very much for this.

December 1st

Finally received my Buwizz 3.0. The mold quality is really top-notch, the Buwizz clips well and around the usb-c there's close to no mechanical play at all. Although I owned the (great) buwizz 2.0, still impressed frankly, that was meticulously designed, you must have had many mold iterations to achieve this I think :) but it was worth it seeing the result. The LED are BRIGHT too Just tried turning it on, will need more time to really test since it's for a MOC but build quality wise, good job. The final target is to have an embedded small raspberry pi connected to the buwizz 3 so will wait for the protocol/potential new firmwares to go public like you said. Also received the buwizz motors, never had the original lego RC motors to compare but the flatter design is great, less space wasted (I think the original take 1/2 stud more ?), we'll see, especially thermal-wise how it fares.

I asked that question shortly after my preorder but when would we be able to have the Bluetooth protocol specifications for Buwizz 3.0. ? Rather than using a controller my objective is to have a small Rpi board managing the buwizz 3.0 :), if you have new information on this it would be cool. Also, should people want to develop their own controller or project it would really be useful. Btw, I think it’s better to drop support of lego bluetooth product in the app like they did, I mean it’s time not spent where it matters because they keep changing the bluetooth API with their firmware, starting to think it’s on purpose… my order was just shipped by the way, can’t wait to test it

Received the update too finally :). It's a small-sized company doing something challenging, so for me it's more than OK the delays, but maybe next time settle the initial shipping dates when you have more confidence ;) as other might not be too comprehensive and realistic about how hard it actually is to manage the development and logistics (especially in COVID-era and global shortages of electronic components and shipping slots. Kudos I have ordered 2 and 2 motors. Very satisfied with Buwizz 2.0 so no reason to doubt the 3.0 :). When the bluetooth API information is done if you could please share them

Thanks for the mail heads up regarding B3.0, happy to see it’s on schedule and working well ! - regarding the battery are they custom made or it’s something standard we can easily find online ? - Also, any chance we’ll see the bluetooth specifications so that we can interface Buwizz 3.0 with any app when the product releases end of this month :) ? The app is great of course but for STEM/programming it would be cool to have those specs

Probably a misprinted 0° deg reference or a technical issue with the motor...

Current APK is < 250mb I believe (I'm on iOS) My humble opinion on the matter would be that either the application has simply many images/animations or that it is not a "native" Android application but rather a hybrid application like Ionic or React Native, those "hybrid" applications often have a memory space overhead You can absolutely have smaller APK for applications, but once you have complex GUIs lots of ondisk images and animations etc that's slightly harder.

Anyone (with some knowledge in software development) can maintain a controller thanks to open, widely-used protocols like BLE. For a specific application yes it will depend but if you are only preoccupied of not being able to drive your creation using Bluetooth, that won’t happen..that’s the beauty of standard protocols and open-source softwares.

Third party softwares are managed by..third parties which aren't necessarily a for profit corporation, some want to release their work for free and did the controller app for fun/hobby maybe other will want to put a price on it, but not all controllers apps need to be purchased. If you have the bluetooth protocol (which we have since it's reverse engineered) anyone can do an app, that's the beauty of it. As the the maintenance and knowledge, it is not a complex task we are talking about, once something goes open-source the knowledge is shared and many people with a developer background (or just people passionated about software development, as a hobby) can basically adapt the application and maintain it. I understand your point of view but I don't share your pessimism, for me the C+ Lego app is just one way of using my C+/PU parts, IF it stopped getting updates (highly doubt so, but in the long-term future) I can still use or create a thirdparty app to use them.

Bluetooth is a standard protocol you already have third party apps to control your lego should lego stop the support on devices. I don't think we are surrounded with phones that have IR sensors should that IR remote break like your sonos (and it's not really premium quality), also I think you can use a bluetooth controller directly or indirectly in some cases rather than a phone. As long as you have device that supports BLE (which is a safe bet) you should be fine. Regarding the official application ,yes it is up to Lego, It's still compatible with devices iOS 10.0+ to this day which covers a lot of iPhones Bluetooth is more future-proof than IR, it is found in nearly all phones/tablet you can buy in the last years.

The C+/PU have an encoder and that is, for me, alone a justification for their existence. I get that you don't need an encoder all the time, that PF is widely available but you can't indefinitely keep backward compatibility, at one point it hinders innovation. For someone wanting to learn (STEM is here the main market), you need an encoder to teach about PID, sensors, stalled torque etc. and just keeping that in a (expensive) mindstorm kit is a shame, better to "democratize" the encoders to all electric motors. Just having a PF with an ON/OFF (unless you use Buwizz 2.0 with different voltages but that's not an official lego set) is not really future proof. Also bluetooth > IR, you can't blame them for wanting to modernize this, a lot of consumers would prefer to control their creation with a better range and on their phone. Regarding the new hub, lot of ports which is great and mostly it can do the job but my main problem..and this is why I'd chose the raspberry pi (in a homemade lego case)/PC + buwizz 3.0 combination is the processing powers and capabilities. They could have went with a SPIKE hub for simple tasks (resource-wise) like micropython, scratch etc. and put a more powerful hub/OS in the mindstorms that would allow web-services, more advanced programs etc to run better, the current mindstorm hub has a 100Mhz CPU I believe, more powerful ARM CPU aren't that expensive to be honest, so the current hub is 'officially' stuck with micropython/scratch and no wifi. I discovered mindstorms in engineering schools so they are used in advanced courses and a shame they did not upgrade it with more modern SoC that would be in phase with today's education : IoT, voice recognition, machine-learning, APIs, machine vision, sensor fusions...

I did the same thing but not with A.I., just merging GPIO/i2C sensors a web-service + lego technic engines in a vehicles using buwizz and BLE. I had a raspberry pi 3A (so no ML on that hardware ^^, maybe tensorflow lite ) in the lego technic vehicle with a lipo battery for the pi. The pi was connected to my 2 buwizz, using the protocol @imurvai reverse engineered. At the time I wanted to train and poc around Node.js so did the vehicle using it, it was more powerful than I thought, Google V8 engine is amazing, but the JS BLE library was not good. I could try and find the code if you need, you have all the characteristics etc, but the github link post here has absolutely everything you need normally ! Will upgrade the vehicle once Buwizz 3.0 is out, with Python3 and maybe some AI this time :). According to @Zerobricks they plan to release the BLE protocol/characteristics for Buwizz 3.0, so hopefully no reverse-engineering will be needed, especially since this time there are way more functionalities.

Thank you for your answers :).