FTA:
--
“Our robot had many features that still haven’t been duplicated today,” Fowler recalled. “We had an autonomous robot that could self navigate, plug itself in to recharge, speak, sing, play music, recite poetry, recognize commands and a bunch of other features. We programmed it to act as an alarm clock, security guard, entertainer and tutor for the kids. It was very advanced.
But when people asked us ‘What can it do?’ and we would list its capabilities, they would always scoff.
--
We see this cycle repeat over and over again in almost every area of tech, and so rarely do we learn from it. What Mr. Fowler failed to grasp was that people weren't interested in buying a robot that had 'features', they wanted one that provided utility.
At the end of the day their lack of utility made these robots toys - very expensive, klunky, difficult to use, and unreliable toys.
Heed me children - when building something, focus on utility over cool features. Solve a problem, and make it easy to use and the world will beat a path to your doorstep.
I had some very accurate voice recognition software on my Windows 3.1 Tandy in 1995 or so. I also played VR games around that same time. The tech has been minimally viable for a very long time. It's the user experience that's evolved and frankly it hasn't evolved all that much.
If anything it's evolved backwards, this 80s bot was doing all those things on its own, without the cloud. And yet now on-device compute isn't enough and everything requires cloud proce$$ing
Speech recognition has gotten dramatically better since 1995. It now works with cheap mics, without per-speaker training, across the room, in noisy environments, with large vocabularies and low error rates.
The problem with this article is that all of this tech is not a binary yes/no feature, but rather work with varying levels of success in various situations.
It also says ridiculous things like "Unlike today’s conversational interfaces, the Gemini took voice commands in an english-like programming language called VOCOL." as if that is a good thing, compared to these interfaces being usable in ambiguous natural language.
I was using an early version of Dragon by Nuance which is still active. It used a cheap commodity microphone and hardware that, while expensive at the time, was maybe 10% as powerful as a typical modern phone.
And having built a public-space voice system very recently, I can tell you that it's still really difficult to pull off and definitely requires specialty AV equipment and professional installation.
The problem then as now is that if it's not 100% accurate, you still need to be doing manual correction which is very time-consuming. And typing was never really a broken system that needed help.
I wish they would at least get to the next half step. Put clothes in the laundry machine, it washes and dries them in the same machine, then buzzes to say it's done. Unfortunately that seems to be a very niche appliance that maybe shows up in some RVs.
You can get some OK ones, but they're expensive. Miele makes the one I have and I'm completely happy with it (comes with full remote control from an app as well).
Does anyone sell a dishwasher where the racks lift out to put away in a cupboard: so you can fill a rack, lift the cleaned rack out the washer and put it away, put the dirty rack of dishes in the washer to clean; rinse and repeat.
A lot of commercial dishwashers have multiple trays - they're integrated into a counter and have two guillotine doors that lift up, allow a new unwashed tray to be slid in one side, while the previous washed one slides out the other side.
If you're washing hundreds or thousands of dishes for lunch or dinner rush, this assembly line is the way to do it.
Dishwashers don't currently handle solid waste well; instead of composting the non-liquid material it's all flushed down the drain, which is both bad for the drain (that grease adds up!) and a loss of the composting that would occur when you manually scrape off a dish.
I was in kindergarten in 1990, and my school had robotics books describing some of these home robots, but they were never very realistic about their actual capabilities. Being a young child, my imagination ran wild.
The Computer History Museum in Mountain View actually has a number of home robots on display, which was cool because it was the first (and last?) time that I could see them in person.
When I was in kindergarten (maybe first grade; not sure I remember which), I called Carnegie Mellon's main line and was redirected to their Robotics Institute where I asked when I could expect to buy a regular robot. I may have told them I'd have to grow up a bit and help them speed things along when the answer wasn't "soon." I also called Learjet around the same time. They even sent me a set of photos and a sales prospectus my parents still have. Though they were less than enthused when they realized I gave them our address and had to explain why a sale wasn't likely to the salesperson who followed up.
There's something truly inspiring to a lot of kids about what engineers can build. Inspiring to me as an adult, too.
This is the source of much of our collective imagination about robots.
And all the dystopian movie plots about robots overthrowing humans and killing them all. I think that is completely projection - somebody is thinking "well, if I were a robot, I'd kill everybody." It really says a hell of a lot more about humans than robots.
For robots to take over the world or kill all humans, first they'd have to give a shit. They'd have to perceive themselves as oppressed or something. And have some motivations, akin to emotions, driving the behavior. Good luck on coding that up.
somebody is thinking "well, if I were a robot, I'd kill everybody."
Most of these, from the earliest, are various allegories about responses to exploitation and oppression not people thinking how they’d want to kill everybody if they were a robot.
IIRC, the word "Robot" itself comes from the Czech play "Rossum's Universal Robots[0]" and translates to slave, so the metaphor of class struggle was always there.
It translates to 'worker', & the play came out shortly before the russian revolution, so it's clear that the subtext of the term in its original use is not merely a criticism of slavery but also a marxist criticism of the exploitation of the proletariat under industrial capitalism.
(Important to note that, just like in Do Androids Dream of Electric Sheep & Blade Runner, the 'robots' in R.U.R. are not mechanical devices but instead genetically engineered synthetic humans. That helps the metaphor a great deal: rather than being a pure 'other' whose behavior might be seen as a design error, it's simply a different breed or race of humans with different heredity that makes them convenient for othering, along with some argument about hereditary traits that supposedly make them more suited to hard labor.)
In all slavic languages, 'robot' is the base of the verb 'to work'. So, by the same means by which some people have decided to translate 'robot' as 'slave', it would be more accurate to translate it as 'worker'.
Asimov's Three Laws don't even work in the fictional setting. A common theme of his robot stories is humans investigating strange or harmful behavior of robots, that ultimately turns out to be a consequence of the laws as the robot understood them. They're good storytelling, but bad real-life ethics.
I never forgot the promised robots. They're ingrained as part of my childhood, a dream of mechanical assistants that were a hybrid of the 1950s futurism combined with the promise of a digital tomorrow. Instead we have been sold cheap spying devices that have more in common with the ubiquitous microphones of 1984 sweetened with the droll dialogue of the talking wall screens of Farenheit 451. Ohh well the future still lives in our dreams and those we choose to build with sweat and grime.
It was an interesting time to be building robots. And most of them had all of their functions running on the equivalent of an Arduino. (generally an 8080 or Z80 though with more RAM and no FLASH). You can read through the archives of comp.robotics.misc for some of the excitement of the time.
Generally, the inability to have end effectors (hands of some sort) that could do things was a real killer. None of these robots could do anything with vision, few had real localization, and about the best they could do was move from where they were to approximately where you wanted them to be without bumping into too many things or getting stuck.
The real robots of the 80s, that did useful work, and liberated an entire class of people from menial domestic labour were the dishwasher, the washing machine, the dryer, and the vacuum cleaner. But, alas, that is not sexy.
Compared to them, these things, and Alexa are toys.
And the machines mentioned in the article sold thousands of units at most.
The theme of the article is that the creators of these machines made a big bet on the idea that novelty and imagination alone would carry expensive, underpowered machines to market success, and that even when those machines were actually technically impressive, those bets were wrong.
The HERO was able to stay on the market for so long only because Heath was making their real money on electronics kits. The RB5X was able to stay on the market for so long because the overhead was close to zero. Androbot & Arctec folded almost immediately, because the TOPO was a $400 RC car and the Gemini was a $9000 tech demo with no arms.
A business plan that depends upon upper middle class frivolity to the tune of thousands of dollars per family is unlikely to succeed. Everybody in the article learned that the hard way.
theoldrobots is a great resource. I didn't cite it directly because I followed it back to primary sources in many cases, but I leaned on it heavily while writing the article. It seems to be the only site on personal robots with anywhere near the breadth and depth of information necessary for even casual research on the personal robot industry as a whole.
It's not too surprising. The first of these were produced in the immediate aftermath of Star Wars' success, & the entire personal robot boom occurred during the original theatrical releases of the trilogy. Plus, a trash can on wheels is one of the easiest mobile robot designs to get working.
It's irrational, but I'd like to tinker around with a HERO robot, too. I say irrational because it's possible to have vastly more capability with a Raspberry Pi-based modern bot, at 10% of the price. But there's something alluring about the golden era "old iron" computers (pdp8/11) and 1stgen micros (8088/6502), tube amps over solid-state modeling amps, and robots. In a sense it's amazing they (we?) were able to accomplish so much with so little resource.
But also, in the same way that cabinet art portrayed 80's video games as graphical masterpieces even when the graphics themselves were little more than blocks, perhaps it is the hopefulness in the face of failure to deliver on the promises of the early robots that makes them so charming.
I hope to live long enough that someone looks at a Boston Dynamics littledog and says "oh, look at that old thing. It's so primitive!"
Talking about "the good ol' days", with mass ammount of resourses we have today, it seems that "we" don't try to make things as efficient as possible, since we dont have to.
Author here, making a quick correction about the RB5X.
When I wrote the article, I believed it was still produced by the original company. I later learned that while the same company produced the RB5X for about 20 years after its initial release, eventually the name & the remaining stock was sold off, twice.
They kept afloat despite only having produced a handful of bots because the stock of parts was kept in somebody's garage & ever few years when an order came it, the gang got back together to build one.
If you've got an interest in these things & aren't confident about your ability to build one from scratch, the RB5X seems like a good choice: help them clean out their garage. If you are confident in your ability to build one from scratch, I recommend downloading the code, designs, & instructions for the Gemini from Mike Fowler's site -- the Gemini is more capable than even newer robots on the market at a price comparable to what they originally asked for, & the price of components can only have gone down since.
Wow! $1800 for the arm subassembly. (IMO, an essential component. If you're going to have a 'bot it must be able to fetch a beer.)
That strikes me as a tough price point to meet. Perhaps I'm jaded after having seen numerous 3-4-5 axis arms on thingiverse & hackaday being 3d-printed and built with servos & stepper motors for a couple hundred bucks or less.
The RB5X incorporates infrared sensing, ultrasound sonar, 8 sensors/bumpers, voice synthesizer, and an optional 5 axis armature that can lift a full pound. The RB5X can play interactive games with up to 8 people. Programs can be written and downloaded from any computer--Apple Macintosh, DOS, Windows, or Linux/Unix.
I remember HERO from my university robotics club, some 10 years ago. Punching in hex codes for machine instructions was so fun, especially since that encouraged us to drop the computers and code on paper.
Nevertheless, someone made an assembler, figured out the code upload interface and future instructions were uploaded using a mini jack and a sound card.
> Alexa’s Interface is treated as revolutionary
> ...it had a lot more in common with an Amazon Echo than a robot in an old sci-fi film
While yeah, these old robots could in theory do more than an echo watch some of the videos of them in operation (eg: https://youtu.be/oibgyqdXfJM).
Just take the voice on that youtube video, while impressive for the 80's is no comparison to Alexa's voice. It is very slow and has a thick robot accent. Alexa has a much more dynamic, fluid voice that is much easier to understand.
And yeah, I'm sure some of these things could be commanded by voice, I imagine that you have to be extremely precise with what you say and how you say it. I imagine the error rate is way higher than Alexa (which in my experience is pretty high itself).
I dunno where I'm going with this comment, but the interface matters. It's easy to claim "they were doing all this shit back in the 80's" and it might be technically true, but how they were doing it didn't have nearly as much polish as modern stuff. When you have a 4mhz processor and 64k of ram, there is only so much polish you can add to an 80's robot. The echo dot has a quad core processor and 4gb of ram -- not to mention wifi, bluetooth and gps (https://www.digikey.com/en/maker/blogs/2017/amazon-echo-dot-...). You can add a lot of polish to a hardware stack like that.
Of course, even with all that processing power, I'm still only about 50% successful turning the lights on at home. We still have a long ways to go....
Yeah. I wanted to foreground the way that ignorance of history blurs the line between iteration & revolution.
Actual revolutionary advances are rare. We haven't had any in the tech industry since the 1970s -- when declarative, functional, object oriented, and structured procedural programming were all formally defined, asymmetric encryption was invented, the first working hypertext systems were implemented, neural nets were first used for image processing, large-scale internetworking & heterogeneous distributed computing were pioneered, and we got all of the graphical UI concepts and models that are in use in mainstream software today.
Everything since the late 1970s has been either slow iteration on 1970s ideas or limited to academic/experimental systems with limited influence.
The BigTrak was my first exposure to programming, as well. I was so distressed when my family moved and it was lost/discarded. I used to work out path programs for it to drive from my bedroom, down the hall to my brother's room, and back, ferrying baseball cards or comic books or such.
For those who aren't aware, BigTrak was an electronic toy from the late 1970s, a treaded tank-like toy car with a membrane keyboard on top. The user could program in a path as a series of steps (forward, forward, right turn, forward, make sound, stop.) and I think it had a gripper claw on the front. (memory is a fleeting thing).
As a kid I thought the whole Tomy line[0] was awesome. I lusted after an Omnibot and Omnibot 2000, but never got one. (I did get a Verbot though.) My imagination still makes me think they'd be awesome, even if I knew they were mostly remote controlled with weak motors, butd damn, seeing a an actual demo of them is heartbreaking.
>The UWCC Biped robot in action at the 1990 Robot Olympics in Glasgow. This robot is a product of Cardiff University. It consists of a pair of life- size aluminium legs. Pneumatic actuators at the joints enable this robot to walk with a rolling gait. Both feet are equipped with strain guages to measure the rolling motion, with a feedback mechanism to prevent the robot falling over.
For kids of the 80s, the Sears catalog was one of the greatest things to peruse. Here's an example of some of the less expensive "robots" that were being offered up in 1986:
I'll never forget the home robot trend. As a CMU frosh, I worked in the "Household Robotics Lab" and we played around with some of those robots. Nobody really knew what they were doing nor why. But we had a lot of fun and had some cool toys to play with. The Raibert Hopper was in the lab next store. That was a serious robot - and seriously loud.
I was barely aware of it. I had (and still have, it doesn't work) an Armatron, and read about some of these robots in books in my high school library and I was fascinated by them. Not smart enough to actually do anything with that fascination, but still... robots.
There was one that had six legs and looked like a viral sheath that I thought looked really cool, but I can't find a picture of for reference. I think it actually appeared on a news program or something.
As a nerdy kid, I was exposed to all this personal robot hype (particularly marketing material from Nolan Bushnell's Androbot) via pop-science library books -- all of them somehow staying in circulation 15-20 years after nearly all the companies mentioned went out of business in '85-'86. Looking for information about the machines in those half-remembered books is what led me to write the article.
I think it would be really easy to miss all of that if you weren't specifically interested in robots growing up, or if you grew out of robot books aimed at children before those books had gotten into the library system. A lot of them were probably published after the machines (aside from the HERO & RB5X) were no longer available, because the publishing pipeline is about 18 months & Androbot went from launch to bankruptcy in about two years.
One of the highlights from my childhood was being served by a Coca-Cola cobot at a diner, refilled my drink for me. It was soon after the Winter Olympics where we got to see it ice skate.
Like all trends, it was localized. You had to live in the right place at the right time.
Me too. It wasn't much of a "trend", there were a few robots available that didn't do much other than slowly move about. I remember the robot "butler". Utterly useless.
Marvin the depressed robot would have been a lauded improvement.
My parents failed to fork over the cash for that (but they did get me the very first Mac, which blew up my 12 year old mind so hard that it probably began my career in software dev, so... I think I won anyway)
But when people asked us ‘What can it do?’ and we would list its capabilities, they would always scoff. --
We see this cycle repeat over and over again in almost every area of tech, and so rarely do we learn from it. What Mr. Fowler failed to grasp was that people weren't interested in buying a robot that had 'features', they wanted one that provided utility.
At the end of the day their lack of utility made these robots toys - very expensive, klunky, difficult to use, and unreliable toys.
Heed me children - when building something, focus on utility over cool features. Solve a problem, and make it easy to use and the world will beat a path to your doorstep.