Mr Hobbles

Just to point out, in this context share of the market is in terms of revenue, not in terms of number of participants. It could be that only 1% of the audience is kidults and 99% are children, but the kidults are spending a lot more. This is to say that there's an argument to be made that by just focusing on where the money is, we risk forgetting our core (and arguably most morally important) demographic. Thankfully I haven't noticed this too mucg yet with regards to Lego's products. For every $700 6,000 piece set there's an excellent looking $20 set for kids. See the latest Lego City wave. But it's a slippery slope.

For me, yes it is. It's not about the battery itself, it's about the device. Phones, laptops, etc, I expect to last ~5 years, as that's the lifetime I expect before I find myself wanting the need to upgrade to something new. Once I upgrade, I have little to no desire to go back to the old one - it serves no purpose that the new one doesn't. With Lego on the other hand, as I mentioned, I'm still using 30+ year old Lego electronics alongside new Lego electronics. Control Lab drives my 9v trains. Powered Up controls my Bluetooth trains. And WeDo/Power Functions controls my sensors, point motors, lights, etc. All controlled and orchestrated from a Raspberry Pi. If I want to use my Smart Brick 30 years from now and the battery is gone, then what? I've written modern programming libraries that enables control of Lego electronics, ancient and new, and I'd love for Smart Play to be a part of the portfolio of devices we can use 30 years from now. https://github.com/nathankellenicki/node-poweredup https://github.com/nathankellenicki/node-controllab/ https://github.com/nathankellenicki/node-wedo https://github.com/nathankellenicki/node-coral https://github.com/nathankellenicki/node-toypad etc

Which is decent, but disappointing. For example I still use 30 year old Dacta Control Lab's on my Lego layout, along with Powered Up, and Power Functions. We probably won't be able to use these in 30 years, and reviving them will be extremely difficult. Don't get me wrong, I understand that they've done it in the pursuit of miniaturisation. But the cost (in terms of a short life) is high.

I’m agreed. I’m pretty sure that if it can be “app controlled”, it won’t be LWP3. I recently put some effort into reverse engineering the Lego Education Science motors, sensors, and remote, and it’s an entirely new protocol. Mind you, I’d say it’s a better protocol - it’s simpler than LWP3. Lego has stated the Smart Bricks have BLE, though in what capacity, we don’t know. We know they can have their firmware updated over BLE, but whether we can control the lights or sound, or read sensor values, who knows. I’m also interested in finding out if the new “BrickNet” mesh capability of the Smart Bricks is related to how the Lego Education Science components work, as it sounds very similar to their own communications capability achieved through BLE broadcasts. My hope is we can communicate with the Smart Bricks through BLE just as we can with Powered Up and Education Science (incidentally, named “Coral” internally within LEGO) But regardless, if Lego is willing to put in the effort, it could be an exciting new ecosystem. Only time will tell.

I love the possibilities of this, but execution is gonna be critical. I love the direction and distance sensing capabilities of the Smart Bricks. The X-Wing vs TIE fighter play demo they did with two kids shooting at each other and registering hits was very cool. Similarly the demo with the racecars crossing the finish line and the winner being registered was cool also. I can see many possibilities - a train that can go end to end on a track by putting smart bricks at each end and reversing direction. Or a police car with a smart brick in it (with flashing blue lights), and a robbers car with a smart brick, and kids chasing them around. I think these things are fundamentally so simple that the end state HAS to be to get the cost of them down such that every kid has a handful of them in their toy box. Then whenever they buy a new Lego set, even if it isn't a Smart Play set, it can come with a Smart Tile, and the build can have space for the Smart Brick, such that anyone can add Smart Play functionality to any Lego set with components they already own. The Smart Tiles are no doubt incredibly cheap. I believe this is already rumored with the second wave of Star Wars sets - Smart Play sets but with no Smart Brick included. My biggest observation so far is that the sound quality is incredibly bad. I know why it is - they've prioritised using synthesised real-time sounds over playback of pre-recorded sounds. They take up less space and its easier to store the sound pattern on the Smart Tile/Smart Minifig itself, as opposed to requiring a firmware update for new sounds. But, it's just not good quality sound. I _hope_ that there some inbuilt storage to unlock the possibility of using pre-recorded sounds in the future.

I made another video showing many things working together, this time with text to describe what's happening.

So, out of the box, with no laptop in the mix, the behaviours are fixed. If they're in the same group, the left joystick on a controller will control the speed of the left wheel on a double motor, and the right joystiq will control the speed of the right wheel on a double motor. If there's also a single motor in the group, then the left joystick will turn the motor on and off - no speed control. Similarly with the color sensor, there are set behaviours on how the motors will react to colors. However if you use a laptop to "code" the devices from the code.legoeducation.com site, then it's like Powered Up or SPIKE - you can program anything to happen, ie. have a robot do a dance when you move a stick up on the controller. Taking it a step further, if you program it using traditional code (ie. using node-coral that I published above), you can make anything - have the lights in your living room turn on if you use the controller. :) In terms of where the logic is run, the devices are dumb - the logic runs on your laptop. The devices receive commands from your laptop and report back sensor data to your laptop. So if you're using the code.education.com site, then it's in your browser.

That’s correct, and in fact sadly, it looks like Pybricks won’t be compatible, as (like the Technic Control+ Move Hub) Lego have started signing the firmware. Mind you, they aren’t programmable, but they are controllable, just like Powered Up. I doubt it’s possible out of the box to have a Powered Up hub join a Coral group, but it may be if you put Pybricks on that Powered Up hub. I’m not too familiar with Pybricks capabilities, but as long as it can broadcast certain services/characteristics/manufacturer data, it’s theoretically possible. It would be an interesting experiment!

Hey folks — I’ve been reverse engineering the new LEGO Education Science components. Much like Powered Up, the possibilities are excellent - provided you’re comfortable getting technical. (Note: According to their code, LEGO refers to this product range internally as “Coral”.) How It Works Each device (controller, motors, colour sensor) has built-in Bluetooth and an NFC reader. When you power a device on, you wave one of the coloured cards next to it, and it “joins” that group. Despite the cards being different colours, grouping is actually handled via NFC, not colour. Each card has a unique NFC ID, even if multiple cards share the same colour. In a classroom setting, this means you can hand out identical-coloured cards (for example, multiple green cards) and still have independent groups, as each card represents a distinct NFC identity. Interestingly, the devices never connect directly to each other. Instead, they broadcast their state using Bluetooth Low Energy advertisements. Other devices listen for broadcasts by the same group and react accordingly. For example: Move the left stick on the controller The controller broadcasts its state The left motor sees the broadcast and responds It’s a surprisingly elegant and robust way of grouping and coordinating devices. Coding Tools LEGO provides a web-based coding environment at https://code.legoeducation.com/ Using Chrome, you can connect directly to the devices and program them via a Scratch-like interface. The blocks will feel familiar to anyone who has used SPIKE Essentials, SPIKE Prime, or Powered Up. Building a Node.js Library My initial plan was to sniff the Bluetooth connection between the Code website and the devices. It turns out there was a much simpler approach: just read the JavaScript! :D While the code is obfuscated, it’s trivial to run through a deobfuscator. Doing so revealed all the Bluetooth commands. Interestingly, the wire-level protocol is far simpler than Powered Up — very lightweight and straightforward. From this, I built a Node.js library in TypeScript, called node-coral, which allows you to control the devices directly using JavaScript or TypeScript. https://github.com/nathankellenicki/node-coral/ Putting It All Together As an experiment, I connected Coral, WeDo 1.0, and a Dacta Control Lab, and got them all talking to each other — interoperability spanning 30+ years of LEGO hardware! (I love this stuff lol) This setup uses libraries I’d previously written, node-wedo and node-controllab. I’ve updated both of these over the past couple of days to bring them up to date. https://github.com/nathankellenicki/node-wedo https://github.com/nathankellenicki/node-controllab Here's the setup: A LEGO Education WeDo (1.0) Hub with a Power Functions Medium Motor A LEGO Dacta Control Lab with a Power Functions Medium Motor and a 9v Touch Sensor A LEGO Education Science (Coral) Single Motor A LEGO Education Science (Coral) Controller The left stick on the Coral Controller controls the Power Functions Motor connected to WeDo 1.0. The right stick on the Coral Controller controls the Power Functions Motor connected to the Control Lab. The 9v Touch Sensor turns the Coral Single Motor on and off. Here’s a video, along with an architecture diagram showing how everything is connected. Hopefully some of you find this interesting - if only it wasn't so prohibitively expensive to normal people (even schools will find this expensive!)

The Pikachu looks a lot like the smart brick Mario. I dunno, I expected it to look different. Since I’ve heard they’re using the brick inside an X-Wing I thought it’d be shaped less like a character and more like a large normal brick. But, as has already been mentioned, that is clearly not the 18+ Pikachu. That’s a different set, a play set.

Lego has been experimenting with 3D printed elements for a few years now, specifically for elements that have motion built into them. I have a few of them. This particular 3D printed train has moving wheels, wheel linkages, and the smoke goes up and down when the train is pushed along and the wheels turn. This isn't multiple pieces that were put together, it was _printed_ together. I do agreed it's an odd choice to include in this set though - it seems like something a more hardcore Lego fan would appreciate than the mass market Lego Christmas set purchaser. I can't help but think how many of these they've printed, they must have dozens of printers going 24/7!

My theory is that at extremely low speeds the PWM duty cycle isn't long enough to actually move the motor. It then realises this (by virtue of the fact that it hasn't detected rotation) and applies a longer duty cycle, which makes it move, but at a higher speed than desired. It then detects it's moving too fast (again, through rotation having moved to far) and reduces the PWM again, causing it to not move once again. Repeat this cycle - leading to jerky motion. I'm thinking that both my motors have slightly different properties (ie. manufacturing variance?) that means that the required PWM duty cycle to start rotating differs between the two.

Thanks for that! Our code is similar, though my button mapping acts a little differently. Interesting that you have 45 as your minimum speed. Perhaps the video was playing tricks but I thought your locomotive was running slower than that. Incidentally, I tried with a second motor of mine. With the original motor, around 30 is the point where movement isn't "jerky". With the second motor, 20 is doable without jerkiness. I wonder if I run-in in the motors for 15-30 mins then slow speed performance will improve. Ps. I was wondering how the heck you fit front glass in with the wire going through the frame, but then I realised, you'd tucked it under. Great little design!

Love the performance. Can I ask what your code for Pybricks looks like? I've attempted similar with motor.run() but movement is incredibly jerky for me below 70. It looks like you've managed to get yours smoother. Great design!

Lego just posted this wallpaper on the Lego rewards site, that can only be found through the new legobuilds instagram. It's pretty cool, and features all the lego modular buildings to date, for the 20th anniversary of Lego modulars (I think they're about 2 years too early?!) (Also, further proof that Lego considers Market Street part of the modular buildings line. :D )

Ah yes, right. Yes, it has very basic LWP3 support - mainly for the specific VM commands used by the Control+ app. While the hub reports a suite of motors and sensors, the LWP3 implementation is woefully buggy and incomplete. Actually similar to the Boost Move Hub years ago before they updated the hub. Additionally, TLG made the hub required signed firmware, making it so Pybricks can’t be loaded onto the hub. I hope they push out some updates for the Technic Move Hub, otherwise it can’t really be used as a proper LWP3 hub for scenarios other than an RC car.

For the new Lego Education Science components?! I know Pybricks has LWP3 support, it's had it for a long time. I'm saying I don't know if the new components have LWP3 support. Ps. I'm the author of node-poweredup on GitHub, I'm very familiar with LWP3. :) I just hope the new components are LWP3 compatible.

Unfortunately I'm skeptical that Pybricks will be able to work on these. The most recent hub (The Technic Move Hub, included in the Porsche 42176) requires signed firmware, to which only Lego have the signing keys. The Pybricks team have said they will not be able to support Pybricks on it. If that is the case with these new components, then Pybricks may be out of the question also. I hope for at least LWP3 protocol support, so we can control them remotely. EDIT: I just saw you hope that Pybricks can talk to them remotely. Yes, I hope for this also - but that requires LWP3. Fingers crossed.

A couple of new pictures from an event at Lego HQ, taken from Boone Langston's Instagram. https://imgur.com/a/HRjhk6e What we can see is the pairing light and USB-C charger on the controller, which suggests an internal battery, instead of replaceable batteries like on the existing Powered UP remote. Along with a better view of the dual motor component, which seems to be 10x6x3, not including the sticky outy rotaty bits. Additionally, this picture is weird: https://imgur.com/a/KcKyQ7o Why are they holding the connector card up to the motor? Can it "read" it somehow?

In one of the higher resolution pictures there are regulatory markings on the side of the motor suggesting it has a battery. From the details, the new components are: * Controller * Single motor * Double motor * Color sensor My theory is that every component has its own battery, and components talk wirelessly. So the controller can control the motors, no wires anywhere, and everything is charged via USB-C. They also detail that "connector cards" are included - these seem to be color coded from the pictures. I wonder if these are used to control the color sensor. I wonder if everything uses LPF2 LWP to talk to each other and is therefore technically Powered Up compatible.

Hi @Toastie - I use this version of nqc which comes with USB tower support, without a requirement for a seperate driver install (at least on macOS, which is what I use). You may find it useful. https://github.com/BrickBot/nqc

I agree that the locomotive isn't great, but by and large I love this set. Also, the Crocodile wouldn't look bad hauling the Orient Express. ;)

It's very nice looking, but far too far off minifigure scale for my liking. Also non-standard track gauge and non-motorisability = hard pass. The Harry Potter crowd is sure to love it though.

I ordered the 1950's Diner and the Modular Construction Site at 10:00am on the dot. They were the two I really wanted. Ordered the Chalet later because...why not. I wish Studgate wasn't so ugly, I'd really like to support a train station...

Man, I'm on the fence whether to get this or not. On one hand, I want to support more trains, and I really like the small train that comes with it. On the other hand, I really dislike the station itself. It doesn't appeal to me in the slightest. Decisions decisions.