Solving The Mystery Of The Maillardet Automaton

A two month long project that fulfills an 8 year long quest to know once and for all what the original writing instrument was for the famous Maillardet Automaton, which is housed at the Franklin Institute here in Philadelphia. The machine was created in the 1790's, almost certainly as a collaboration between the shops of Jaquet-Droz and Maillardet ~ with the majority of the mechanism designed and manufactured under Jaquet-Droz and the programming (a very large bank of 3-axis cams) created by Maillardet. The automaton is essentially an elaborate 3 dimensional pantograph, and so I use a simple 2-D pantograph in a proof of concept project to replicate the way the Maillardet Automaton originally made its elaborate renderings with very high crisp detail using varying line widths. I conclude this one as Q.E.D. - Quod Erat Demonstrandum!
Franklin Institute Feedback - www.fi.edu/contact
My original blog "Reverse Engineering the Maillardet Automaton" -
www.frantone.com/designwriting...
Join Team FranLab!!!! Become a patron and help support my KZbin Channel on Patreon: / frantone
#hugo #lost #mystery
- Music by Fran Blanche -
Fran on Twitter - / contourcorsets
Fran's Science Blog - www.frantone.com/designwriting...
FranArt Website - www.contourcorsets.com

[@chuckgenco]

This video is an example of what I love most about your channel. We get a glimpse into how engineers, with limited methods, solved extremely complicate problems with cleverness and a kind of brute force will. It reminds me of the Arthur C. Clarke quote “Any sufficiently advanced technology is indistinguishable from magic.” Except it is kind of in reverse. It's not the advanced technology that seems like magic, it is these early solutions by amazing engineers of the past.

[@TheErador]

All electromechanical engineering seems like magic to me. The amount of complexity they were able to model mechanically is phenomenal, given the lengths they would have to go to for certain features. Thousands of gears etc etc

[@gabotron94]

"Any sufficiently advanced engineer is indistinguishable from a wizard"?

[@senilyDeluxe]

When it was explained to me how 1960s era gigahertz frequency counters work using electronics that only work up to a few tens of megahertz by doing lots of fancy maths*, it dawned to me: Any sufficiently antiquated technology is indistinguishable from magic. (I think that was on CuriousMarc's channel) *) and so many more devices that do stuff that was considered impossible

[@ferreusveritas]

Any insufficiently advanced technology is distinguishable from magic.

[@NoJusticeNoPeace]

I remember reading an article about the founder of Connection Machines (who make mainframes), and it mentions that in university he constructed an analog computer -- essentially an automaton -- in his dorm room out of fishing line which was capable of playing a perfect game of Tic Tac Toe.

[@trespire]

The work you did on this automatron has historical significance. These early machines are the blueprint for many principals in industrial automation that heraled early mass production and the Industrial Revolution. This is huge ! Well done.

[@ElectricGears]

What seems almost more impressive is that the builder had to make an entire other machine to make the programing disks. Basically a mechanical means of converting Cartesian to polar coordinates with built in non-linear transformations due to the mechanics of the arm. I imagine it was done in a similar manner as your panto-graph setup. I wonder if the arm is back drive-able which would let you produce a disk with the automaton by replacing the follower with a cutter? I would assume there would be clues in the mechanism because you would need to account for backlash in the opposite direction as the writing. If I was tasked to build this, I think I would make a scaled-up replica of the arm and use scaled up engravings as you did with the panto-graph. How plausible would it be to install a new set of disks? That would seem like a very useful feature the designer would be crazy not to include. It would be pretty cool to map the mechanism in 3D which should allow the drawing-to-disk profile function to be derived and new cams made on an CNC machine for new drawings. I suppose you could also mechanically scan the cam profile.

[@FranLab]

Not really with the mechanism as it exists today but without the clockwork it could have been possible - but making the cams was a complex and likely multistage process, and the cam bank is extremely difficult to remove.

[@ElectricGears]

@@FranLab Darn. I just think it would be really funny to have a fully restored, extremely elegant, Victorian era clockwork mechanism drawing dick butts.

[@brucesheplan696]

Reading your comment gave me this idea on how this could have been made: - Build a 3-axis pantograph where each axis has a separate stylus/engraver - Draw your design as you record each axis independently ( imagine like the tracing of an EKG or a seismograph) - ideally each line would be on the same piece of paper so you dont have to worry about synchronization - wrap each pattern onto a disc - use the discs to drive each axis on the automata I think the hardest part would be the wrapping onto the discs as there may be some distortion when you do this, one solution would be to engrave the discs directly then cut them to the engravings It would really interesting to see someone attempt to replicate the automaton.

[@Lucius1958]

@@brucesheplan696 I wonder, also, whether the original inscriptions might have been copperplate engravings, which could be traced by the pantograph onto the cam blanks for all three axes. It would require some pretty complicated mechanisms to synchronize the pantograph with the rotation of the cams, though...

[@lwilton]

@@Lucius1958 Remember that routers didn't exist back then, and even a drill press was likely to be hand-turned. You couldn't have just had a blank cam plate and milled it to correct depth as everything was moving. About the best you most likely could have done is to scribe a fine line on the side of the cam disk, and then take it out of the machine and hand-file it in a vise.

[@emiliaolfelt6370]

I've always found these machines super fascinating, and listening to you talk about it? 10/10 You're cool as hell Fran.

[@EEVblog2]

WOW! This is amazing! Great work Fran.

[@rbrooks2007]

I love automata. The BBC showed a fantastic documentary about them. What blew my mind was clock makers hand making screws with a thread diameter smaller than a human hair.

[@BenState]

try electronics

[@rbrooks2007]

@@BenState Production is done with computers. They don't carve silicon or make the chips individually on a workbench by hand.

[@stormtrooper88]

@@BenState i like discovering the easter egg messages some engineers have written and left for others to discover. they truly are humorous sometimes. i have made a few myself when designing circuit boards years ago. i am sure they got a few chuckles from fellow engineers. i mean who doesn't like smiley faces. i just could't resistor doing it. =o)

[@normalizedaudio2481]

Franklin is so lucky to have Fran.

[@randomhuman1965]

We are All lucky to have Fran

[@KeritechElectronics]

I wish they did. But no... again, some petty poly-ticks keep them from employing one of the biggest brains on the net!

[@wildfire405]

I am brand new. I've been watching your videos for a total of 4 minutes or so. You are INCREDIBLE. Thank you. I don't know what it is yet, but you're fun, and the tech and science information you're sharing flows right up my nerdy, maker, sciency arteries right into my heart. Fantastic!

[@Seele2015au]

Fran, if the brush hypothesis is correct, then there would be a few points of interest. First it would need to be a "fountain brush" with a controlled internal ink supply, something that's quite difficult to achieve for it is without a nib and feed. Then the ink would need to be a pure dye type, with no pigment or material to form a polymerized matrix. Then the bristles would need to be very bouncy so as to recover from any bending and springs back into its original position precisely: bristles like this do exist such as kolinsky but they do not last very long due to breakage. William K Rockman's two US patents, issued when he was living in NYC, tackled these issues directly, but a wet brush-head exposed to air for any period of time is certain to ruin the brush very quickly. If that be the case, the automaton would have been one that required the writing instrument prepared and installed before each demonstration by someone who understood the operation of the fountain brush in question.

[@hanzzarkov7690]

I cant imagine in a thousand years having such amazing penmanship. I'm convinced that even if i spent a decade with a clock maker, they couldnt refine my particular 'style' of chicken scratch.

[@flymypg]

Splendid! I knew nothing about brush pens, and am delighted to see how the line quality (ink flow and line width) appeared to be consistent over time. Most detectives would be satisfied with solving the crime. No, not our Fran. She has to replicate the crime in every detail! QED, in its purest form.

[@dtsdigitalden5023]

The game Syberia piqued my imagination when I played it years ago, and I never realised such complex clockwork-robots actually existed until your video. This Maillardet Automaton takes the imagination to unusual places, particularly with respect to intense amount of tinkering and refinement (and the time it must have taken) that must have been required. All the hours you spent (with modern tech at your side) gives insight into the immense amount of time it must have taken to create the original automaton. Fascinating video, and seeing your work gave remarkable insight into what an achievement the automaton was.

[@stevechilde2319]

To get z-axis control you could use a conical stylus so that it would move down in a wider part of the groove. The width of the line would control the depth. You would need to have a smooth engraving to work on as some others have said, but one way might be to use a cnc router and code in the depth for each part of the line according to its width. With both of these working (?) you may be able to lock up the hinge on the stylus arm so that the z movement of the stylus would control the z movement of the pen, and effectively apply more pressure where the line needs to be wider. This might (?) reduce the skill required. Another thought about the lines crossing - with metal plates for the templates and some registration pins or a frame it would not be difficult to substitute templates so that no template would have too many difficult crossings like the fine lines crossing the thick oval. Great project, really got me thinking!

[@rubikmonat6589]

Basically a record groove. I wonder if you can untangle the crossovers with some asymmetric left and right gain in the text-parallel direction.

[@TeeDohJackson]

I've been obsessed with Maillardet's automaton my entire life!!! The fact that you got backstage passes to the greatest show on earth and got to come face to face with it makes me so excited for the future of this magical mechanical masterpiece 😃

[@slide6strings]

Quite amazing. I saw the Automaton as a kid in the early 60s. It's wonderful to see it in action again. Your commitment, persistence, and creative problem solving are truly impressive. Well done!!! Hope your hands have recovered from the vibrations.

[@robertburns2415]

You are a most skilled and patient person. Congratulations on your outstanding achievements.

[@cliveradvan3414]

And not a single 3D printed object insite (yay)! Pure genius Fran.

[@zombie-process7025]

Fran, I'm impressed but absolutely not surprised they had you work on this restoration. Your skill and expertise is world-class.

[@paulcooper8818]

Very interesting. Got to admit I thought you were going make a pen holder for a 3D printer. And then use the printer's Z axis motion with the brush pen to render the artwork's varying line width. So for awhile I was thinking "What's this all about!" Thank for making this video.

[@argus5323]

this is the first video i have seen of yours found it on my home page... the thumbnail suckered me in to watching it. (good job) i just have to say "You go Girl" i truly admire your patience to do this task let alone make a video of it. i cant imagine how much time was put in to making the original Automaton, things like this just blow me away. I truly believe the previous generations were way smarter than the people of today because of things like this. Thank you so much for everything you done to make this possible.

[@davidnuckols8151]

You have just showed us a tremendous project that must have taken several weeks to accomplish. Bravo Fran you are awesome.

[@joelmurphy9369]

Thank you so much for telling this story. So clearly reasoned and fascinating! the footage of the 'naked' Maillardet is an intimate and very special sight to see!

[@tactileslut]

Given the careful balance of pen and arm, do you think the weight and floppy inertia of the sleeves would have messed with the rendered image?

[@joelmurphy9369]

​@@tactileslut i doubt it. the mechanical parts are too strong and positively driven to cause mere linen to restrict its moves. This thing was transported many miles, in pre-modern transport, and ran and ran and ran many times to perpetrate the lie of machine intelligence. ... ahem, I mean to say that we don't even know if machines have lives, let alone intelligence.... perhaps a very small intelligence? like the kind that keeps the climate control consistent? This machine looks very strong, but also looks very precarious. Does it require absolute level installation? Lots of linear shaft movements.... But still, if it will run through the moves, you can expect the result will be within margin of error. Many automaumata machines use this same machine-to-gestural maneuver. Not only could it run for a long time still, it could be copied. Scaled. Modified.

[@mrbrown6421]

7.15.2023 Wow. That was awesome. I learned a lot about the joys of proof of concept and dedication. Well done! Thank you.

[@peterpegrem9837]

Esto Fidelis is my old school motto. To understand the future you must be faithful to the past. You are a light on that Fran.

[@paulkinzer7661]

Layers and layers of excellence! I learned so much in just this one video of yours, and you have so many others with totally different topics. Your skills and knowledge make me feel so fortunate to have found your channel!

[@abundantharmony]

Automatons have fascinated me since I was a child. The engineering behind them blows the doors off of the mechanical engineering of today. A truly lost art. Glad you are digging in! Wonderful video.

[@David-is1jn]

Videos like this one are why I’ve been a patron of yours for as long as I have~ love this video =]

[@Vermilicious]

How fortunate you are to have been able to interact so closely with that machine. I'm sure only a select few has had the privilege.

[@boscorner]

I love seeing obsessive passion projects like this. Another absolutely fantastic video from fran

[@erictuffelmire6826]

These automaton from centuries ago are very fascinating.

[@CARLiCON]

creating this machine: amazing, creating this machine in the 1790s: mindblowing..it's a treasure & a miracle that it made it to the Franklin Museum- thanks for sharing

[@Chrisamic]

Brush pen. Oh, I think you're exactly correct in working out what the automaton was set up to use. That's brilliant. Now I've worked out that my stationer actually has the Tombow N15 black brush pen in stock, and they are only seven bucks. I've just got to have one. I know what I'm doing first thing tomorrow morning. I can't get a blackwing pencil anywhere in this country to save my life, but brush pens in sooooo many colours are there to play with. I'll start with black and see what happens from there... Regarding 3D printing... yeah it's hard to start with. Learning 3D modelling software is an investment (same as any CAD really). Once you have the skills it's quick and easy though. I just got a piece of test gear (watch pressure testing unit) that turned up with a smashed thumbscrew for the pressure lid. I got a $15 (6.5%) refund and an apology for poor packing from the supplier, and one hour later I had a new thumbscrew with the original threaded brass insert installed on the unit. The casual observer would not even notice the difference to the twin part on the other side (but my part is better ;-) I'm using FreeCAD which has a steep learning curve and lots of idiosyncrasies, but I can get the job done pretty quickly these days and I enjoy the experience. I don't count the printing as "work", though it seems frustrating to wait ten or twelve hours for a print sometimes. It's not that bad - I just start it in the morning and get on with other things. Currently I'm printing tool holders for my new watchmaking bench.

[@eDoc2020]

She already has a 3D printer...

[@Chrisamic]

@@eDoc2020 Yeah I know. Fran doesn't like modelling, which is why I was talking about that. I was trying to be encouraging.

[@AveragePicker]

What an amazing look at this. Thanks so much for sharing!

[@performa9523]

Absolutely incredible! Mad props Fran!

[@sfdntk]

I'm very annoyed the algorithm didn't surface this for me when it was first published, this is exactly the sort of content I come to KZbin for, and I've been subscribed to you for years! Fantastic video Fran, thank you for everything you do.

[@apexmcboob5161]

That was fascinating and I have to add your perseverance at this was inspiring! One of your best videos, thank-you.

[@MyRealDelusion]

you know what I always see in amazing people is the uncanny ability to put your head down and just keep going the drive to do something to the end separates great people. bravo.

[@DifficultNerd]

This is an utterly fascinating story. Thank you for sharing with us, Fran!

[@yttkuar]

Amazing work Fran. What tenacity and patience.

[@dmeemd7787]

Friggin INCREDIBLE work!!

[@RedHeadForester]

Fascinating video. Thank you Fran!

[@MacEstelle]

Fran this is an amazing video! Your detective work, and approach to replicate this very complex technology YOUR own way is nothing short of inspiring. Many thanks for sharing.

[@maccamcdermott5528]

Bravo, Fran! Your dedication is absolutely awe inspiring.

[@dmeemd7787]

This channel is so great, love it from day one! This is AWESOME you are working on this!! So awesome!

[@J.DeLaPoer]

Excellent work, and a fascinating video! As someone who's both a watchmaker, and into calligraphy & vintage pen collecting as hobbies, it is absolutely and instantly obvious that a brush would've been the only type of pen/nib that could possibly produce the variable line widths (AKA "shading" in proper writing parlance) the automaton originally did. Any conventional flexible nib regardless if it's gold, steel, quill; antique or new, simply will not write when held vertically; nor when moved across paper in an upward direction. All nib/pen types other than brushes produce shading via direct downward pressure to spread the nib open, while being pulled down the paper toward oneself. That's the only way they function: Upward strokes and/or never removing the nib off the paper, as the automaton writes, would only rip the paper and damage the nib itself. So it is indeed a brush or nothing for proper and original results here. The other pen types that will "work" are ballpoints, rollerballs, and stylographics certainly. However as you say, _none_ of those will produce anything but uniform lines. *I would hope if they ever start exhibiting this automaton again, they'd take your suggestions into account rather than slowly damaging it over time by forcing it to write with a standard ballpoint.* A ballpoint is a workaround that will A) not produce any shading eg; line width variation as it properly should, and B) causes wear & tear on the delicate mechanism. It's forcing it to apply significant downward pressure on a hard, inflexible ballpoint; when it was originally designed for a brush which would be almost literally weightless in terms of pressure required eg; putting zero strain on the mechanism. It pains me to see incredible, one-of-a-kind clockwork mechanisms of such genius getting slowly ruined by simple ignorance...

[@AerialTheShamen]

They could make that ball pen spring loaded and/or use a felt pen to reduce risk of damage.

[@soundenguy]

Great work as always Fran, love your dedication to your projects and knowledge.

[@darold1966]

Thank you for sharing your work.

[@WCM1945]

Fran is probably the most multi-talented person I have ever seen!

[@RufianEmbozado]

What a beautiful clever work! Thanks.

[@bostedtap8399]

She's alive, fantastic engineering across the centuries, original concept is astounding. Thanks Fran for leaving us this legacy. Sincerely wishes from the UK. John.

[@robertheckman9813]

Wonderful video Fran. Thank you for putting in all the hours of work.

[@profmoustache5217]

Mind-blowing work, Congratulations Fran!! 👏

[@PierreaSweedieCat]

Simply AMAZINGLY clever!

[@vremster]

Fascinating! I love the combination of art and engineering.

[@CajunReaper95]

I swear I use to watch your videos several years ago, glad I find your channel again!

[@jimmy_jamesjams_a_lot4171]

I was just at a hardware store with your Origin router on demonstration. It indeed does not look like an easy thing to become acclimated with. It seems to me an odd sort of tool to adapt for use with CAD renderings. That automaton is absolutely mesmerizing. THANKS AGAIN SO MUCH FRAN, for doing the research that you do to bring us your content that is all educational and very entertaining! You are an invaluable lector and a very inspirational, top notch human being!!

[@jts3339]

Fascinating. One of your best videos ever!

[@jameslewis7620]

Wow! This is amazing Fran! 🤩

[@gratefulot360]

Fascinating! Thank you!

[@KeritechElectronics]

What a fine early example of numeric control! Overshadows my beloved Monotype text composition system for sure.

[@robertgift]

Wonderful! Thank you. Howonderful that you gained such access!

[@jonathanvanier]

Amazing work, Fran! A significant contribution to history and curatorship!

[@fepatton]

That’s a brilliant solution! My first thought was a flexy nib too, but as soon as you mentioned the vertical orientation, I could see that wasn’t it. I admit I’m surprised that all you could get was 15 minutes with the original!

[@kimbledunster]

Fascinating stuff! A very plausible theory.

[@rippenbug]

Absolutely awesome content, always amazed at your knowledge!

[@DullPoints]

I love your workbench! Great explanations as always. :D

[@cyndicorinne]

This is an enthralling look at some great engineering work, yours and the automaton creator’s.

[@keaphotoscom]

And that Fran is why you should be a national treasure. Thanks for all your work.

[@AgentOrange96]

With a more sophisticated CNC system, it might be possible to adjust the depth of the pantograph template to control the Z height. This would be similar to the grooves in a record that encode both left and right channels. This would require programming in the line width and using some sophisticated coding probably to turn that into a height level. So it'd not be super simple.

[@paulmoir4452]

Yeah, just use a 3d printer: enough precision, no need for much force. Programmable with G-code. There's a gross negligence of "thinking with 3d printers" in this video. Adapt a pantograph not using parts you make with your time rather than the 3d printer's time, develop useless skills with said adapted machine rather than learning some modelling or programming.

[@winterwatson6811]

fran needed to prove a concept, not make art. your approach may also work, but she’s the one who actually did the thing. i hope you enjoy your projects and your workflow.

[@hadibq]

very satisfying work!! great job Fran!

[@jazzerbyte]

Interesting analysis of a complex machine along with the proof of concept!

[@MegaPeedee]

Amazing. It had me enthralled. I have used pentagraphs in my art but nothing like this. I have been fascinated for years with the automatons and you have filled in the blanks for me. Thank you.

[@submarineradioman5535]

Fran - I remember this from childhood trips to the institute! What a privilege to be able to work with this machine!

[@marchawkinson2548]

Thank you for your work and commentary! I watched this because I remembered this sort of thing from the movie Hugo.

[@fibberscloset498]

Absolutely fascinating. Thanks, Fran.

[@marc.levinson]

This is so insanely cool and complicated for such a small detail.

[@Trishlicious]

I love your workbench, that is lovely.

[@TheNightshadePrince]

Professor Trelawney explaining how a clockpunk boy's writing works has made my life complete. :)

[@podfuk]

Brilliant Fran, thanks for sharing!

[@redbeacon4871]

Frans voice is so relaxing in addition to what her audience learns and absorbs in her awesome videos

[@iwasinnamuknow]

What an absolutely fascinating machine.

[@Traderjoe]

You are brilliant Fran! I really appreciate how you yielded to make the pen holder manually rather than 3D printing something. That technology is great, but I am pretty sure the bulk of your audience probably can relate to your method here than going to CAD and plotting it out through that method

[@supralapsarian]

Fascinating. Thank you for sharing the details and fruits of your process.

[@terryolsson4145]

Wow Fran, this video was facinating. You're amazing.

[@michaelboty2346]

Your respect for the maker/s of this machine must have increased with this whole exercise. Thank you.

[@luluscohen]

I never realized how much I needed to know the answer to this mystery until Fran solved it.

[@robertrobert7924]

So nice to see you again Fran. I went on a field/bus trip to the Franklin Institute when I was in Junior High School. I was fascinated and hooked on museums from then on. Living in Baltimore, I took many trips to the Smithsonian Museums and National Zoo. Conscription into the US Army ended my college career in 1968 where I majored in Anthropology. Thru some miracle I was hired to work by The National Museum of the American Indian as it was being built on the Mall. I worked there for 4 years from 2000 to 2004. It was the best and lowest paying job I had ever had.

[@KirkIsmay]

Thanks, Fran! I enjoy your videos and exploration of the past!

[@FranLab]

$$$$ Thank YOU Kirk $$$$ Kiiiiirk!

[@TentoesMe]

That's beautiful, Fran.

[@lo-firobotboy7112]

I'm totally with you regarding 3D printing. It cracks me up seeing KZbinrs print square blocks or simple standoffs etc. fabricating simple shapes and tools is still faster and easier with good ol' fashioned hand tools and simple materials.

[@azz710]

Simply amazing.

[@wesleysale1052]

Outstanding! You cover so many useful topics that interest a lot of people.

[@0756rocketman]

Hey Fran, thanks for the wonderful work that went into this video.

[@Unemployedrobots]

Fascinating! Dang Fran!

[@ryanreedgibson]

This is really cool. I am glad I came across it.

[@jjones503]

Never seen your videos. But I have to say this intrigued me in a way no other video has ever done so before.

[@drewd1726]

Such a great video. Thanks for being you Fran. Waving from Oregon. 👋

[@kennymanchester]

Well done Fran! I sincerely hope that you get to personally validate your work on the actual machine. As a caligrapher, and watercolor painter, I believe you have absolutely nailed it!

[@ColinTimmins]

This video was really cool to see. I love this kind of stuff!

[@AlanCanon2222]

I'd missed the FranLab theme song (been watching a lot of 16mm uploads).