Humanoid Robot community dying off?
Humanoid Robot community dying off?
by patkangu
Mon Jan 27, 2014 5:00 am
Hey Guys. Been noticing that on this forum and on many others, (Trossen Robotics etc.) the amount of activity has been dwindling and slowly grinding to a halt. Is this because there are no new developments in the robot industry or is the community dying off? No new small humanoids have been released recently and I've seen less and less activity from companies like Hitec, Kysho etc.
Re: Humanoid Robot community dying off?
by PaulL
Sat Mar 15, 2014 11:55 pm
I figured I'd wait a while and see who replied to your post here... Some "tough love" to follow in this post...
Hitec discontinued the Robonova-1 - I'd guess they didn't sell as many as they thought they would. As far as I know, Robotis is still doing its thing (though, only sales numbers can tell the true tale there). These small humanoids haven't had significant innovation in a while, but just whose fault is that - the manufacturers, or ours? Both? If someone IS innovating, it sure would be nice to know (well, I'm doing my thing, but that will continue with or without Robosavvy! Can't recall the last time someone replied to my posts under Home Built!). Even Robosavvy seems to be focused on 3D printing more than 'bots these days.
I can't say for certain why this particular board is dying off, but I can make a guess or two:
First and foremost, forums in general are dying out, casualties of more immediate forms of gratification - FB, Twitter, name-your-favorite-site-here. Few seem to want to spend time at a keyboard, better to twiddle a smartphone with your thumbs for a few seconds here and there, I guess.
Other than the general forum die-off on the 'net:
The focus of the main page isn't the forums and doesn't look a thing like the forums, other than the color scheme - it doesn't even have a list of recent posts on the front page, and "Forum" is on the far left or far bottom of the list, depending on how you're viewing the pages here. On a desktop, forums launch in a new window, though that basically turns the main page into a "pop under" of sorts if you're just here for the forums.
The forums themselves aren't as visually appealing to me - I've said it before and I'll say it again - I don't like the "new" look of the board. The new site is, to me, too simplistic, too bland. It doesn't have enough visual complexity, and I think visual complexity (ex, barage of bot info - products, posts, pages, stuff that is interesting) is kind of expected by roboticists - not in a chaotic way, but more in a useful way. The colors are rather bland, and I'm not a big fan of the new logo (sorry to those that put in the effort) - it just doesn't represent a site that, to me anyway, used to be the go-to place for info, particularly for small humanoids. When you visited Robosavvy.com in the old days, you had it all right in front of you - recent posts, products, info. It had the look of a mature, evolved site. The old logo just screamed "you're at THE site for hardcore robotics enthusiasts". The Kondo in the logo was a better representation for the site than even the RN-1 (and for those that read my posts, you know I'm an RN-1 fan), it always had more of a DIY feel, a bit more of "imagine I have a face" - keyword, "imagine". The new logo says to me, "I look like an owl" - it just doesn't click to me, doesn't really speak to me of a site for robotics enthusiasts. Again, I do apologize to those that have put forth the effort.
Historically, we had the old board, then forum spammers become a significant problem, which didn't help things. Then, there was the board / site switch. Meanwhile, social media has become even more of a focus, though I personally don't care for those sites.
More recently, we used to have posts from Robosavvy of various efforts in the way of their own robots, but even those have faded. (btw, RS, what ever happened to the contest winner's bot? I haven't seen much of anything about that!).
Personally, I don't want a bot that looks like a doll for a child, I want a 'bot that looks like all he needs is a bit of scale to take on the world - and yet is functionally powerful enough to take on a smaller piece of the world. I'd love to have a full sized bot, but practically, we're not there yet, way too expensive the bigger you go.
Cost is an issue, always was, always will be - these bots aren't cheap - 16, 20 servos alone will set you way back. Then, there are the controllers. There's no migration path from the factory-provided controllers, no easy way to upgrade to, say, a PicoITX SBC.
Then, there's software: off-the-shelf bots have custom software, and some manufacturers try their best to provide extensible systems, but still, you're stuck with their software unless you are much, MUCH more adventurous, and even then, you can't grow beyond the provided hardware.
Roboard was a good start at PC-based bot controllers, but I personally stepped back from it over audio problems. Since then, there are faster and faster boards available (but lack bot-specific hardware). I walked away from Roboard due to audio related problems, problems with full duplex audio and CPU saturation. It has cost me a ton of time, looking for alternatives, and ultimately buying a different SBC and building custom boards to do the bot-specific stuff - not to mention that the SBC I bought is considerably larger than the Roboard and required a torso redesign just to fit it in.
In the end, the only ones that can say for certain what happened to visitors is Robosavvy - I'm sure they have logs of when their forums started to die off...
I have to say, I do wonder what the former "hardcore" posters are doing these days.
Hitec discontinued the Robonova-1 - I'd guess they didn't sell as many as they thought they would. As far as I know, Robotis is still doing its thing (though, only sales numbers can tell the true tale there). These small humanoids haven't had significant innovation in a while, but just whose fault is that - the manufacturers, or ours? Both? If someone IS innovating, it sure would be nice to know (well, I'm doing my thing, but that will continue with or without Robosavvy! Can't recall the last time someone replied to my posts under Home Built!). Even Robosavvy seems to be focused on 3D printing more than 'bots these days.
I can't say for certain why this particular board is dying off, but I can make a guess or two:
First and foremost, forums in general are dying out, casualties of more immediate forms of gratification - FB, Twitter, name-your-favorite-site-here. Few seem to want to spend time at a keyboard, better to twiddle a smartphone with your thumbs for a few seconds here and there, I guess.
Other than the general forum die-off on the 'net:
The focus of the main page isn't the forums and doesn't look a thing like the forums, other than the color scheme - it doesn't even have a list of recent posts on the front page, and "Forum" is on the far left or far bottom of the list, depending on how you're viewing the pages here. On a desktop, forums launch in a new window, though that basically turns the main page into a "pop under" of sorts if you're just here for the forums.
The forums themselves aren't as visually appealing to me - I've said it before and I'll say it again - I don't like the "new" look of the board. The new site is, to me, too simplistic, too bland. It doesn't have enough visual complexity, and I think visual complexity (ex, barage of bot info - products, posts, pages, stuff that is interesting) is kind of expected by roboticists - not in a chaotic way, but more in a useful way. The colors are rather bland, and I'm not a big fan of the new logo (sorry to those that put in the effort) - it just doesn't represent a site that, to me anyway, used to be the go-to place for info, particularly for small humanoids. When you visited Robosavvy.com in the old days, you had it all right in front of you - recent posts, products, info. It had the look of a mature, evolved site. The old logo just screamed "you're at THE site for hardcore robotics enthusiasts". The Kondo in the logo was a better representation for the site than even the RN-1 (and for those that read my posts, you know I'm an RN-1 fan), it always had more of a DIY feel, a bit more of "imagine I have a face" - keyword, "imagine". The new logo says to me, "I look like an owl" - it just doesn't click to me, doesn't really speak to me of a site for robotics enthusiasts. Again, I do apologize to those that have put forth the effort.
Historically, we had the old board, then forum spammers become a significant problem, which didn't help things. Then, there was the board / site switch. Meanwhile, social media has become even more of a focus, though I personally don't care for those sites.
More recently, we used to have posts from Robosavvy of various efforts in the way of their own robots, but even those have faded. (btw, RS, what ever happened to the contest winner's bot? I haven't seen much of anything about that!).
Personally, I don't want a bot that looks like a doll for a child, I want a 'bot that looks like all he needs is a bit of scale to take on the world - and yet is functionally powerful enough to take on a smaller piece of the world. I'd love to have a full sized bot, but practically, we're not there yet, way too expensive the bigger you go.
Cost is an issue, always was, always will be - these bots aren't cheap - 16, 20 servos alone will set you way back. Then, there are the controllers. There's no migration path from the factory-provided controllers, no easy way to upgrade to, say, a PicoITX SBC.
Then, there's software: off-the-shelf bots have custom software, and some manufacturers try their best to provide extensible systems, but still, you're stuck with their software unless you are much, MUCH more adventurous, and even then, you can't grow beyond the provided hardware.
Roboard was a good start at PC-based bot controllers, but I personally stepped back from it over audio problems. Since then, there are faster and faster boards available (but lack bot-specific hardware). I walked away from Roboard due to audio related problems, problems with full duplex audio and CPU saturation. It has cost me a ton of time, looking for alternatives, and ultimately buying a different SBC and building custom boards to do the bot-specific stuff - not to mention that the SBC I bought is considerably larger than the Roboard and required a torso redesign just to fit it in.
In the end, the only ones that can say for certain what happened to visitors is Robosavvy - I'm sure they have logs of when their forums started to die off...
I have to say, I do wonder what the former "hardcore" posters are doing these days.
Re: Humanoid Robot community dying off?
by limor
Tue Mar 25, 2014 7:45 pm
The forum is mainly for the hobby community
people here on the forum show off their creativity and we've seen some amazing stuff over the years.
humanoid robots in their multi-servo format and this forum have now been around for 10 years.
The hobby community is usually attracted by state of the art and cutting edge stuff
Therefore the classic Kondo style humanoid robot is now gone "mainstream" in the sense that it is now selling more to schools than to hobby
However... I expect some evolution of the humanoid robot hobby sector through introduction of new actuators which in turn will create new visuals that will compel hobbyists to take a fresh look at the domain. (see what happened with quad copters and then with gimbal brushless motors direct-driving the FPV systems - reduced complexity and cutting edge offered to hobbyists and it became a billion $ market within 2 years; gradually eroding from the previously lucrative expensive RC cars and helicopters business)
The RC style servos with 1:200 reduction have a lot of backlash and low back-drivability.
To make a robot that performs like a Boston-dynamics or asimo, we need an actuator with direct drive, big torque and low reduction gear and backdrivable or something like a harmonic drive that has no backlash but is not backdrivable at all.
people here on the forum show off their creativity and we've seen some amazing stuff over the years.
humanoid robots in their multi-servo format and this forum have now been around for 10 years.
The hobby community is usually attracted by state of the art and cutting edge stuff
Therefore the classic Kondo style humanoid robot is now gone "mainstream" in the sense that it is now selling more to schools than to hobby
However... I expect some evolution of the humanoid robot hobby sector through introduction of new actuators which in turn will create new visuals that will compel hobbyists to take a fresh look at the domain. (see what happened with quad copters and then with gimbal brushless motors direct-driving the FPV systems - reduced complexity and cutting edge offered to hobbyists and it became a billion $ market within 2 years; gradually eroding from the previously lucrative expensive RC cars and helicopters business)
The RC style servos with 1:200 reduction have a lot of backlash and low back-drivability.
To make a robot that performs like a Boston-dynamics or asimo, we need an actuator with direct drive, big torque and low reduction gear and backdrivable or something like a harmonic drive that has no backlash but is not backdrivable at all.
