Ask HN: Good programmable robot kit for teens?

51 points by fenier ↗ HN
Hi there,

I know a few teenagers interested in robots, but are not super sold on the blockly drag and drop programming.

Anyone have something slightly more complex, while still being geared toward Teens?

Thank you!

39 comments

[ 4.8 ms ] story [ 77.5 ms ] thread
Maybe Lego Mindstorms Ev3? [0] It's super cool and programmable with UI programmer app and PC/MAC assistance, though still with programming blocks. But I see people managed to use Python for programming it [1]

[0] http://www.lego.com/en-us/mindstorms/?domainredir=mindstorms...

[1] http://bitsandbricks.no/2014/01/19/getting-started-with-pyth...

Also Java via Legos. The mind storms main brick supports booting off the Microsd card without modification into alternative environments. The hardware is more expensive than Adriuno, but it's a lot quicker startup to getting it to do something.
About 8 years ago, I used LeJOS to do Java programming on an NXT brick. It looks like they've updated to support the newer EV3. It's nice to see that they're still in active development.
Playing with Mindstorms+NQC (http://bricxcc.sourceforge.net/nqc/) is one of the main reasons I love programming. I am sure there is an updated version for the latest evolution.

I later got into Handyboard+IC, which is even more powerful/hackable.

Seconded, Mindstorms EV3 are great. Check out ev3dev (www.ev3dev.org) if you want more programming options. You can build some serious robots with this setup, but it's still pretty easy to use.
Looking to get one of these for my nephew during the holidays
Did your kids use it? what was your experience? I took a look and it looks damn complicated - so many parts, how do kids manage to assemble this; Me as a kid would have lost interest fast ...
I don't have first hand experience, just discussed the thing with a friend of mine.
As a kid I had many big Lego sets but not Mindstorm. I always enjoyed building those. Recently, I bought Mindstorm set to play with. I don't see it much different than regular Legos. Only difference is programming part but you can always download pre-built program and just tweak a bit.
EnduranceRobots.com is looking for enthusiasts and hobbiests in robotics and laser industry. We are looking for tech smart people who would like to help us to improve our products and positioning on the markets.

We are still very early startup and can not pay big salaries but we can pay some.

Please have a look our web site: endurancerobots.com youtube channels: http://www.youtube.com/channel/UCzziBor2UytyKZ2RCPjXvCA http://www.youtube.com/channel/UCLVtLSTxwLPs74x1cUoAHzA and our facebook: https://www.facebook.com/Endurance-1399951053640252/

We are very open to all critics: email us your ideas to: gf@endurancerobots.com or skype my: george.fomitchev

Mindstorms EV3. Hands-down, no question about it.
If you have the cash, we use Turtlebots in our labs. They use ROS which can be programmed in C++, Python or LISP. There's a cheaper similar version that also uses ROS called DeepBot, but I haven't had any experience with it. It looks good from the website though and that price is hard to beat.

ROS is great as the code you write for one robot can be used on another one fairly easily. It's also gaining pretty large traction in industry with companies. I find that Raspberry Pi/Arduino route more geared towards hardware. I tried that route myself and lost interest fairly quickly.

[0]http://www.turtlebot.com/

[1]http://www.ros.org/

[2]https://www.autonomous.ai/deepbot-research-robot

Check out FIRST. If you're lucky a local high school or organization will already have a team that you could have them join.

http://www.firstinspires.org/

It's a yearly competition and teams build a robot to compete. That robot is usually pretty big (3x3 ft. at the base, usually?). There's a kit of parts that includes a board, and the programming environment is mostly pre-configured for you.

Next year's build season kicks off early January. The stuff these kids can build in a month is pretty amazing.

Seconded. I grew up doing the program and owe a lot to it. There's different levels for all ages.
I volunteer with our local FTC group - FTC is grades 7-12, FRC is 9-12.

FIRST has really made our lives more difficult, paving the road to hell with good intentions by using android phones instead of microcontrollers, which required all new hardware of pretty poor quality and a Java software stack that some teams are finding difficult. Our local group has been farmed out around the state to provide support much more often than in years past.

That said, FIRST is probably still the best organization for these kids to learn robotics, and I don't know whether FRC is going to suffer the same hardware and software hurdles as FTC.

I was a mentor (mostly doing software) for a couple high schools, and I can also recommend this. Even as a mentor, I learned a lot from helping high school students program their robot.

They provide a nice platform to get you started, but you still have to do a lot of thinking and clever engineering to get it all built.

The teams I was with did welding and created parts on CNC machines, but that's not necessary. Many teams build their robots out of purchased pre-fabricated parts. That's a much better way of doing things when your team is new.

You get the task the robot is supposed to perform and the rules it needs to follow in January, and you have to build a working robot in about six weeks. Everyone is very busy, and there's a lot of things that go wrong as the thing gets built. Everyone learns a lot from the experience.

ComputerCraft turtles on modded minecraft.

If you must spend money on hardware, buy them desktop upgrades, more memory or the fastest processor that'll fit in the socket (may need to upgrade BIOS first...) or a top of the line graphics card.

I think turtles are easy to program, they're in LUA and are a step above blockly type stuff. I kinda like the LUA APIs, they're not ridiculous or anything. They're simple enough that you'll get a chance to build useful larger functions / subroutines.

At least its cheap. If they throw up their hands in frustration at programming their own tunnel mining program, at least you'll find out soon and it'll be cheap. Start with building a robot smart enough to build stairs all the way down to bedrock. Extend to have it place actual stair steps and torches for light, etc.

Note that there's a subtle difference between a future-EE-teen who thinks programming robots is cool and a future-EE-teen who thinks robot hardware is cool (like motor H-bridges and gray code position encoders and sensors and generally melting solder). Obviously the latter is not going to find modded minecraft amusing. There is also the "competition robot" set which usually isn't programmable but amounts to homemade RC cars, you specifically don't want that, just mentioning it as something to look out for and avoid in the market.

https://www.packtpub.com/packt/offers/free-learning

Register, and use this link to add ebooks to your personal library. I've added several books about Arduino and e.g. Python. You have to login each day, and the book can only be claimed that one day. You don't know what book will be free tomorrow, and yesterday's book is gone.

I think the DFRobotshop Rover is a pretty nice starting kit: http://www.robotshop.com/en/catalogsearch/result/?q=dfrobots...

It's really simple and straightforward construction, but it's also really expandable, and really standardized. The main system board is basically an Arduino with a bunch of additional parts like the motor controllers pre-included. You can drop in XBees for wireless control or what-have you, add Arduino shields, etc. The drive system is a very cheap, standard Tamiya kit as well, so it's very easy to repair/replace damaged parts.

And it's an Arduino, which means you have a programming platform that's pretty simple and exploding in popularity right now.

IRobot has a new hackable robot geared towards STEM education. It's basically a stripped down (read: vacuum-less) roomba you can hack on.

www.irobot.com/create

(comment deleted)
Try building from scratch! Super fun.
I'm currently developing a FOSS line of robots aimed at tweens, teens, and everything in between. They are not ready yet (work gets in the way!). I know the marketplace well. Your best bet is to go with a Parallax Boe-Bot with the Arduino instead of the Basic Stamp. The kits are high quality and cost around $150 per kit. They are durable and are extensively documented. One kit per 3 teens is the norm. Anything else will have a bunch of plastic proprietary parts that break and not be as well documented. With teens you want a solid road plan that will keep them interested without overwhelming them.

If you have any questions, feel free to email me (profile). :)

Robot maker here so had to chime in with an obligatory plug :-)

I produce Mirobot which is designed to teach kids about everything from the engineering aspects of how it's built to programming it using a number of different methods. It's a drawing robot so you can also use it to learn about geometry and maths.

It's all open source and easy to program because it connects via WiFi which means it can all be controlled via WebSockets in your browser. There are a bunch of apps (http://apps.mirobot.io) which include some drag and drop type apps but also an app to program it with JS. But because it's all just async JSON over WebSockets (or raw sockets) you can program it in any program you like really.

It's Arduino compatible which also means you can dig in at that level too.

Check it out if you're interested: http://mirobot.io (15% off with code: XMAS15)

I've been involved in robotics for a long time. The thing that I find attractive is that they engage three different engineering disciplines, electrical, mechanical, and software. However, that same diversity can often dissuade a student who is interested but becomes intimidated by the steep learning curve.

The other thing about robotics is that you really need a goal, otherwise people lose interest. The HBRC started the table top challenge which was good for a variety of reasons, 1) it was approachable, 2) it needed only a table top to work, and 3) it had enough complexity (in stages) to keep people challenged. DPRG created a number of challenges as well that were similarly staged.

So first ascertain which of the three disciplines they are most likely to be successful at. And start there. Parallax offers a number of good kits which are easy to program and the Makershed has the Arduino based robots.

Second come up with a goal, an objective that you seek to accomplish. Working toward that will give you the feeling of accomplishment you need to stay interested and feel like you are going somewhere.

Third, decide on your budget and set your scale appropriately. For low budget robotics you can build robots using converted servos for motors that roll around on tables, but for higher budgets you might want to build "Magellen" robots which can navigate around outdoors. If you really want to burn money quickly I suggest starting a battlebots team :-).

Shameless plug, but...

I was frustrated by the lack of affordable, expandable walking robots, so I've been working on one of my own

http://www.redrobotics.co

Not for sale yet, but I'm starting to run workshops locally and the plan is to crowdfund in the early spring with a target price of about £70/$100. It'll work standalone but can hold a Raspberry Pi or Arduino for expandability.

I think legged/armed robots can be a lot more engaging than buggy robots, and open up some interesting avenues for creating motions and thinking about multi degree-of-freedom limbs, but things like Robosapian are just toys, and the $400 hobby robots can get boring quickly (e.g. adding a camera is difficult).

If you have access to a 3D printer there are some interesting open source robot projects at the moment - for example Poppy: https://www.poppy-project.org/ for which you can start by just building a torso