With this robotic arm demo video from 1975 the future really is now. Think about it, there are entire factories full of the descendants of this technology where the human workers simply feed the beast and fix it when it breaks.
We’re pretty impressed by what’s shown off below. Not because we see something we didn’t know was possible, but because the technology was so advanced nearly forty years ago. Here the arm is laying out a wiring harness on a jig. We wonder if using a single color of wire is going to make it a major pain when they add the connectors?
Obviously the mechanics were solid. Time has brought further advances in precision, reduced costs that make robots available for even small factories (often palletizing products is done by a machine similar to this), and improvements in how tasks are programmed. After all, the ability to print a hard copy of the program as a punch tape isn’t quite cutting edge for this decade.
What does that mean for you? If you look hard enough you might be able to find an older generation robot arm to hack on.
[Thanks Amos via EE Times]
Retrotechtacular is a weekly column featuring hacks, technology, and kitsch from ages of yore. Help keep it fresh by sending in your ideas for future installments.
23 thoughts on “Retrotechtacular: 6CH Industrial Robot”
Hi, this guy “Christian S.” has toyed with a Manutec R15 (machineweight 320kg / 15kg lifting / 6 axis)
Amazing isn’t it ?
The tech fan appreciates the modern manufacturing and it’s opportunity
to remove a lot of the body damaging
(I’m that old and have some “damages”)
If you watch some the videos of current robots
You’ll understand why a lot of jobs will never be available again
No matter what part of the world.
Not meaning to be an anti tech comment,
just trying to offer an honest, pragmatic assessment from my perspective is all.
If I didn’t have a like of and interest in ever changing tech, I wouldn’t be reading here
on such a regular basis.
and yeah I think the bots are awesome tools.
I’d love to take one of the robot chair rides someday.
Have they got all the code glitches out yet? XD
those arms & motors just don’t understand that someone
miss-typed a bit and that floor is awfully solid.
While I part ways with Peter Maurin often, I find this from his agrarianism/industrialization easy essay interesting to ponder. ;
4. When all the world
will be looking
for foreign markets.
5. But when every country
you will not have
Could it be that the globalization is great crowd in time will shit in their own nest?
@Mike, you ever seen an aircraft’s wiring harness? 99% of the wires are white.
Also, some car manufacturers switched to monocolor wiring in part or whole some decades ago for both repair job and actual security(when hotwiring was actually a thing).
I would not wager that the wires are white just from that image. the wire loom/binding material is likely white, but the wires themselves are likely color coded. There’s probably an elaborate coloring scheme with attributes such as “main color, line1 color, line1 dashpattern, line2 color, line2 dashpattern” creating thousands of ids.
to some extent the wires would need to be identified. figure them out with a volt meter would be impractical and error prone later.
One end of the harness has the wires already in their correct spots, so the only difficulty is sorting out the other end. Identifying them would be a matter of separating the wires and putting them in an automated measuring jig that scans the wires and swaps their positions accordingly.
Nobody’s going to go through the color codes by eye to plug them in at the connectors one by one. The color coding would be there for later repairs if necessary, but at the manufacturing stage, there’s lots of better ways to identify which wire is which.
Two ways it use to be done was little coloured rings would be slipped onto the wires (think resistor codes), that was replaced by tone tracing.
You plug one end of the harness in to a computer, and it’d transmit a tone down each wire in sequence. You find the loose wire at the other end, connect it up and the computer moves onto the next one.
Lots of companies still do harnesses by hand on a jig like shown.
But I’m pretty sure the circuit they belong to is printed on them in repeating pattern.
We still got 6 similar robots in factory (they move a bit faster, one iv just checked is from 1982). They are like old cars – make sure power is on and everything is lubricated and they will work for another 30 years.
I’ve been programming ABB IRB 6400 robots for some time now. I love all the engineering that has gone into them.
They combine mathematics, mechanics, programming and much more. Everything a Mechanical Engineer should love. The coding on ABB robots is very robust too, it’s high level enough that you can take some pseudo code to production very fast.
I know I’m spewing random useless facts now, BUT, another great innovation of ABB robots is the use of quaternion rigid body rotation and translations. Meaning they don’t have to worry about singularities during it’s movements, additionally it only requires the definition of 4 variables to define its rotation instead of 6 of other robots using homogeneous transform matrices or axis by axis degree definitions.
Man, I freaking love robots!!
Here is an example of one of my robots:
I worked in a small machining/fab shop
(early 80’s) and helped fab up a few lift/griping fixtures. UHMW was a blast to do layout work on.
But typically pneumatics were used to apply griping force.
..along with choosing the types of grip surface material the designer had to set griping force
and sometimes flow velocities
(preventing air cyls from slamming etc)
Sometimes he would be back in the shop
with an actual product/package in order
to tune the rig to task.
sometimes even a change in the box wall stiffness or texture would cause the need to change grip force or shape and size of a grip pad.
forgive the pre-ramble :) but I was curious
how much you run into this and sort of thinking
about the the difference of having the options to alter programing to change the tool control
for slight product line parameters?
Sorry I’ve been up too late and am
probably writing in a terribly ADD manor.
ALL THE TIME! For most consumer production parts, the palletizing/containers are always an afterthought. Most of my projects have been this way, the product is in the pipeline and no one has thought to budget for packaging so I need to be fast and cheap. However, if I do have some foreknowledge of a product, I can plan via CAD and test end effectors and cell layouts before I ever cut aluminum.
Programming is the single easiest and quickest way to add more flexibility to a robot cell. If you cannot program your way out of the problem, I then look at end effector improvements (end effector improvements never stop) to help pack the product while still performing current duties. Lastly, I would look to the cell layout improvements or a justification for another robot.
I am a fan of 80/20 aluminum extrusion and SMC cylinders. That in conjunction with a CNC, you can make some wonderfully beautiful end effectors. I love using aluminum because I can cut some very thin walls 1.5mm-0.5mm which help create an end effector that is light and rigid.
What’s great about the robot programming is that you are forced to think about all the variables. With larger high speed palletizers like the IRB 6400, you HAVE to know your product mass, center of mass, moments of inertia, as well as, your tooling mass and moments. If you don’t account for those you are going to wear out the robot and totally blow your tolerance window. Although the robots are large ~2m tall with 3m reach it has a 0.3mm repeatability, without compensating for your variables, you’re lucky to get 10mm.
I could go on and on about end effector design, best practices for programming, cell layouts, assembly, cnc. Automation is just a fantastic thing to me!
Thanks for your replys (and other peoples too).
My dated background in motorcycles (dirt&street)
Ridden year round & dirt bikes were considered fun in the snow etc.
throw in a range of repair and maintenance
and enough job changes to sound like I’m
120 years old.
Gives me great appreciation for what you say about graceful movements in the coding
and machine life span.
Regarding the UHMW:
I envy having water jet and laser cutting so readily available now. We freehand cut and you’d
need to scribe the layout with a needle (stub of sharpened filler rod) and then wipe the stock with
a small rag wetted with Dykem just to get lines
to cut by.
didn’t get to use much alum in the fixtures then
and carbon steel struct tubing was a always
a bit varied in the interior lumpiness of the weld seam.
We had a couple of excellent repeat clients in packaging itself. i.e. cereal boxes, panty hose inserts, etc, even some mylar film coating apps.
the engineers there actually were concerned
with the slipperyness and how corners broke
Worked at another Co that did press conversions
from solvent based letter press ink to Flexo/water based inks for corrugated industry. They also designed and built some presses
die cutters and so on.
So i got to see even more of the packaging world with them.
sad/embarrassed to say I’ve got one of their “old” PC’s
in front of me and the orginal HD (put aside for a backup) still has auto cad and a few other
not too cheap progs that I’ve just never found a
motivation to learn.
Its too bad ABB robots have a few undesirable nick names in the integration community/industry. Another Broken Bot, Always Be Bootin, plus a few others I can’t recall. Yes, the rapid programming language is quite robust, I’ve always called them a programmers robot, but do you really need to be able to perform trig functions when welding together car bodies or stacking a pallet of bricks? They’re overly complicated robots for maintenance personnel, which is who usually takes these over when I leave an install.
Never heard those nick names before. I agree, they are a programmers robot. The multitude of functions I believe are more or less a relic of older operating systems of ABB. With older robots with floppy drives you have a limit of about 285KB per program, even though the floppy can hold 1.44MB. With those space limitations you are forced to keep the programs compact and limit your positional data. For palletizing programs on the older robots I have modular subroutines which will procedurally generate all positions. For a cell with multiple products going to multiple places it saves a ton of space. The newer OS on ABB has made positional programming trivial because you have gigs of space, where you can import large paths from CAD data, or lazily go to a point, learn, go to a point, learn. In my opinion, the use of procedurally generating points is far more robust and flexible than point by point operation.
We purchase used robots a lot, and as long as the PM schedule is followed the robots will last a long time. Our oldest right now is a M94, so 19 years old. However, I will concede that when the robots are programmed by a rookie, or improperly maintained, or crashed repeatedly, they will have many problems stemming from one issue. I haven’t had a great deal of trouble maintaining or repairing my ABBs from simple fluids change to full rebuilds or calibrations. They are very well thought out machines. Unless, you don’t know what you’re doing, then they are a nightmare.
This is a video from a research project in AI using robotics. They used an analog camera and video processing for object detection and identification. The robot assembles the components.This is from 1973. It’s pure awesome.
The fragment is part of “The Lighthill report” – a discussion that needed to settle the question whether AI was a field of research with a promising future or just a distraction without foreseeable merit. The debate is boring at times but some of the arguments will make you cringe. One of the prominent participants is John McCarthy.
Edit: starts at 00:18 but the link dropped the offset.
This is because google had some performance problems, and youtube is google
It’s pretty standard in industry to use the same colour wire. They are however, numbered – sometimes along the entire length, sometimes just on the ends where they join the connector.
This is an interesting bit of history from a time when most people thought robots were only in science fiction movies. Meanwhile, researchers were initiating some of the technologies we still base our work on today. Robotics typically handle mundane tasks which leaves more challenging work for employees. Rather than eliminating jobs, robotics will create jobs for people with specialized training. It is a great time to encourage your children to take an interest in robotics by joining a Legos club or technology club at school. It could turn out to be a really lucrative career for them in the future.
I believe it was the economist Keynes who was the first to predict a three day workweek…..years ahead of the Jetsons cartoon, lol. Ironically, automation does have the potential to force the city worker back to tend the fields, where he would have the ability to drop out of society, and go back to a more fruitful and rewarding 20 hour workweek like our ancestors enjoyed…..but then, who would line the pockets of the supply siders and money hoarders?
Please be kind and respectful to help make the comments section excellent. (Comment Policy)