In 1958, the American Machine and Foundry (AMF) Thermatool Corporation (later known as AMF Corporation, later acquired by Prab Company of Michigan) initiated an R&D project for a Versatile Transfer Machine, or VERSATRAN, a programmable cylindrical coordinate frame robotic arm designed by Harry Johnson and Veljko Milenkovic. AMF introduced Model 102, a continuous-path transfer device, and Model 212, a point-to-point transfer device, in 1962.
AUTOMATIC HANDLING EQUIPMENT CALLED 'VERSATRAN'.
3243.02 | AUTOMATIC HANDLING EQUIPMENT CALLED 'VERSATRAN'. (1:02:10:00 – 1:05:52:00) 1967
Hatfield, Hertfordshire. Date found in the old record – 23/02/1967.
Various shots of the 'Versatran' – an artificial arm and a hand construction grab which is controlled from large panels. Developed in the USA by American Machine & Foundry Company. The grab is seen picking up a large bobbin and placing it in a box. The control panel can be programmed in advance so the grab can be operated in advance. Demonstration by Mr D C Hall.
It was only in 1967 that the Tokyo Machinery Trading Co. in Japan imports and sells the first industrial robot, a Versatran from AMF, Inc. Britain aquires its first Industrial Robot, a Versatran, in 1967, by Douglas Hall, as seen in the video clip above.
RISE OF THE ROBOTS by George Sullivan 1971
A second industrial robot arrived upon the scene in 1963. Manufactured by a division of AMF Thermatool, Inc., this robot is called the Versatran ( from versatile transfer ). It is characterized by a sturdy horizontal arm coupled to a six-foot vertical steel column which is mounted on a rectangular base.
Although they [Unimate] are different in appearance, the Versatran robot and the Unimate have many similarities. Both can handle objects weighing over 150 pounds. Both are built to last for forty thousand working hours. They sell for about the same price, approximately $25,000 [1971].
Industrial Robots at Work
Industrial robots do work of every imaginable type. They spray-paint automobile engines and spot-weld auto bodies. They stack brick and pluck hot parts from presses and die-casting machines.
What the robot does depends on its program. With the Versatran robot, there are two types of program controls. One is called point-to-point control and is the type used for relatively simple jobs. The other, for more complex tasks, is called continuous-path program control.
When programming a point-to-point control operation, the arm movements and functions to be performed are first drawn on a piece of paper. Then this sequence of "orders" is translated into electronic signals. Short lengths of metal-tipped wires, known as "patch cords," are inserted into the holes of a small, black pegboard, called a "patchboard," to correspond to the written orders.
The programmed patchboard locks into the robot's console panel. The board's contacts connect with memory-storing and command devices known as "potentiometers." Once the potentiometers have been adjusted for the various arm positions in the cycle, the machine is ready to operate. The robot user may own several patchboards, each programmed for a different job.
Programming the Versatran robot for "continuous path" operation is a matter of "teaching" the machine the proper motions to follow. A switch in the console is set for "program." The operator then leads the robot arm through all the motions it will later assume on its own. Gripper commands are also acted out. These signals are automatically recorded on magnetic tapes within the control console. There are fifteen minutes of program time available on each of the two reels of tape the console contains.
The Unimate is programmed in similar fashion—by moving the robot arm through the desired sequence of operation. The sequence registers in the machine's memory unit. Once the robot arm has been "taught" a program, it will follow the prescribed set of operations over and over.
"There's no mystery to programming," says one user. "It doesn't even require a mechanical background, much less a knowledge of electronics."
The job the robot is programmed to do may involve several individual tasks.
See Milenkovic tribute and mention of Versatran development here.
VERSATRAN robot in the 1971 movie "Silent Running"
The "billiard's" playing robot is actually an AMF Versatran industrial robot.
Regarding "Silent Running", for a 1972 movie, the Versatran was still considered a state-of-the-art industrial robot.
Two interchangeable end-effectors are shown, a gripper for loading/depositing billiard balls, and a pneumatic "cue" to strike the ball (below). The standard two-fingered Versatran gripper picks up a B.A.S.E.(tm) 3-fingered gripped to deposit the balls. Another small continuity error in that when picking up the B.A.S.E.(tm) gripper, the 2 pneumatic lines are not attached, but then magically appear in the next shot (see above). Also in the above image is the AMF Versatran name/logo, as well as the controller on the left.
The control panel in the background is a real and actual point-to-point Versatran control panel, used to program the various movie sequences. Although portrayed as "thinking for itself" , this robot would have to be choreographed and programmed via the point-to-point controller.
Google translation (uncorrected – see brief form above):
You toffee robot elephant
Pilgrimage across the country you toffee robot elephant
Caramel absolute throne, occupied by Hisashi Mori milk caramel into a yellow sack after the war. Further around 1955, appeared one after the new caramel flavor. Toffitaipu ones like Britain, which had previously been made from a type of Asotokyande, Toffikyarameru Hisashi Mori was released in 1957 as the first of the so-called pockets (£ 20) is a powerful counter propaganda were characterized. A sack of caramel, it was not designed the figure of the Palace Guards UK. To make several thousand life-size tin signs named after this place we head the country's leading confectionery and went in a big eye-catching campaign.
The same year Hisashi Mori in the press inviting Korakuen amusement park in Tokyo, conducted a naming ceremony and presentation of the world's first robot elephant. Put the real teeth height 2.4 m, length 3.15 m giant elephant with a sophisticated and you are named toffee.
This pilgrimage to the country you toffee, newspapers and radio around the hustle and bustle of the parade was reported extensively in television. Equipped with a 25-horsepower automobile engine, our walking, moving Nosshinosshi neck and nose, a distance and did not put genuine distinction. Visiting 162 cities nationwide, 20,000 km of total stroke, total distance 660 km walking robot elephant, during the visit about a year and raised the great hype effect.
The Elephant appears to have rigid legs with powered wheels. Its head is also larger than other Mechanical Elephants I have seen.
At the debut press conference, Handyman twirled a hoola hoop and wielded a hammer.
Handyman slave station being held up by a G.E. "O" Man.
Handyman was built between 1958-59 at Schnetectady, New York for the joint AEC-USAF Aircraft Nuclear Propulsion Program by Ralph Mosher.
The idea of CAMs originated nine years ago when the General Electric Co. was in need of an especially delicate manipulator to handle experiments with an atomic aircraft engine. Manipulators with the theoretical dexterity to turn screws, fit parts and assemble close-tolerance components were available. But they always turned out to be clumsier and more inefficient than expected. The company asked Mosher, a mechanical engineer in the General Engineering Laboratory, to try to design a manipulator that could handle the task.
"I realized that after a certain point improvements in mechanical dexterity added little to a manipulator's performance," says Mosher. "So I began wondering why a human being is so efficient and a slave robot so awkward. Soon it was obvious that the manipulator's operator was missing what he ordinarily experiences, a sense of feel."
Mosher toyed with several methods of restoring a tactile sense to a manipulator's operator before he hit on force feedback. The idea itself wasn't new; power steering, for example, which became popular in the early 1950s, uses the same principle. But no one had ever applied it to a high performance slave robot. When Mosher did, the difference was dramatic. "We didn't just make a better manipulator," he says. "Adding touch created an entirely new kind of robot."
From Mosher's work came Handyman, a pair of arms with pincer hands sensitive enough to pack eggs, strong enough to crush golf balls, and adroit enough to light a match. The robot proved to be the most effective linking of man to manipulator ever built.
But it did have serious drawbacks. One was the electronic force feedback system, which used sensors in the robot's pincers and arms to pick up and relay stress. The electronic equipment was bulky and far too complex to be practical. In addition, the cost of the servo mechanisms and follower racks was prohibitive for anything but specialized projects. General Electric put Handyman and similar CAMs to work in its own plants. But their complexity and expense nixed plans for continued development.
For a further description on Force and reflective feedback, see post here.
Video Clip - there is a video clip, but, alas, no preview is available. Here's the clip description:
Mechanical Man Clip Description:
Schenectady, N.Y. — Manipulator-Mechanical Man
Cut Story:-Three shots Robot moving arms, 2 shots man operates same, Med Robot removes pipe from tube, Med operator, Med Robot twirling hoop, stops & holds hands together over head. Various shots Manipulator in action.
"Handyman"-Mechanical Man.
(18776) (NXO 2518) 16mm
Story number: 147-183
I found another video clip that has small snippets from the above video clip, but just showing the claw only.
"MM7" on the left, Claus Scholz in the middle and "MM8" on the right.
Scholz made MM7 between 1957-8 and finished with MM9 in 1973, so I believe.
The MM7 Selektor human machine is the development for which the international scientist Scholz-Nauendorff, nicknamed the "Viennese father of robots", is best known. Designed as a means of studying cybernetic movement the MM7, with its feedback stepping switches and visual receptors, is widely regarded as the predecessor of today's industrial robots. Its inventor later concentrated on artificial thought processes in an attempt to motivate follow-up models to act autonomously.
Cybernetic Machine MM7 Selektor, 1961
Claus Christian Scholz-Nauendorff [1915-1992]
Inv # 21918
MM stands for MaschinenMensch i.e. MachineMan.
Popular Mechanics 1964 incorrectly calls this robot MM47, it is MM7.
Younger lady is Scholz daughter, I believe.
Note: MM47 is a typo, should be MM7.
MM7's younger brother MM8 (Contina).
Notice the later modified head on MM8 between the images above and below. The rear head mount and grill-mouth are changed. This is an upgrade of MM8 .
Upgraded MM8.
MM8 is controlled by another 'phantom' or remote control, looking simpler than MM7's phantom.
The combination of 'phantom' and controls and motors makes MM8 much more responsive than MM7. The video clip shows MM8 brushing Mrs. Scholz' hair, near impossible for MM7.
Story
MR. KLAUS SCHOLZ, A VIENNESE INVENTOR AND ENGINEER, HAS DEVELOPED TWO ELECTRICAL ROBOT "SERVANTS" WHICH ARE DESIGNED TO PERFORM VERY EXACTING TASKS.
AS WELL AS ANSWERING THE DOOR AND THE TELEPHONE, THE ROBOTS CAN POUR DRINKS, SHAKE HANDS, HANG UP CLOTHES AND VACUUM CLEAN FLOORS. TO ANSWER THE TELEPHONE, THE ROBOT USES A MINIATURE TAPE RECORDER DEVICE WHICH IS BUILT INTO ITS HEAD.
THE MOVEMENTS OF THE ROBOT ARE CONTROLLED BY COMPLEX INSTRUMENT PANEL IN THE INVENTOR'S WORKSHOP. MR. SCHOLZ HOPES TO DEVELOP HIS MECHANISM EVEN FURTHER SO THAT EVENTUALLY HIS ROBOTS WILL BE VISUALLY ABLE TO RECOGNISE AND REACT ON KNOWN OBJECTS. HE ALSO PLANS TO BUILD A MACHINE WHICH WILL DO ALL HOUSEHOLD CHORES – EVEN THE WASHING-UP.
Reference 2661/64
Can 3734
Source REUTERS
Date original 16 MARCH 1964
Duration 1.49
Technical 16MM/NEG.
16MM/POS.
Subset Reuters TV – RTV Post 1957
Location VIENNA, AUSTRIA
Sound
Colourbw B/W
1. MV FIRST ROBOT AT DOOR 0.05
2. CU "SCHOLZ" NAME PLATE ON OTHER SIDE OF DOOR PAN.. TO ROBOT OPENING DOOR 0.13
3. CU ROBOT MOVES EYES 0.21
4. MV ROBOT CLOSES DOOR 0.27
5. MV KLAUS SCHOLZ AT DESK 0.29
6. CU SCHOLZ AT CONTROLS 0.34
7. MV DITTO 0.36
8. CU VOLTMETER 0.39
9. CU SECOND ROBOT TAKES VISITOR'S HAT 0.51
10. CU FIRST ROBOT 0.53
11. CU SECOND ROBOT 0.59
12. CU SECOND ROBOT'S EYES MOVE 1.04
13. MV FIRST ROBOT WALKS WITH BOTTLE IN HAND 1.11
14. CU FEET MOVING 1.16
15. CU BOTTLE OVER GLASS HELD BY SECOND ROBOT 1.21
16. CU SCHOLZ MANIPULATES CONTROL FOR POURING ACTION 1.25
17. CU BOTTLE TILTS 1.27
18. CU POURS DRINK INTO GLASS 1.36
19. CU SECOND ROBOT RAISES GLASS TOWARDS HEAD 1.44
20. CU FIRST ROBOT 1.49
Another video clip here, thanks to the author of a comment below in locating it.
Stills from clip.
MM7 (and MM8) remotely controlled by a 'phantom'.
Scholz was also experimenting in voice recognition and ….
…pattern recognition.
Claus Scholz.
Translated from Spanish by Google
A creature that is not gossip
Last "robot" built by the Viennese engineer Claus Scholz, who has been engaged for years in this kind of gadgets. Went do remote electronic control, open the door when the doorbell rings, the phone serves, if Mr. or Mrs. are not home, takes message on tape and plays it when convenient. It also handles the vacuum or serving a drink if asked. In short, perfect maid or butler, the days off without pay.
Caption: Taking a wrong message by telephone without names is something that "MM-7" made without difficulty. (Improved image from another magazine)
Caption: The lady who goes to visit her friend is frightened by the disquieting aspect of the "maid". But you will get used to her. (Improved image from another magazine)
(Improved image from another magazine)
Caption: [Below image] Will need to understand that children of the house, if any, are gradually accustomed to a different home as physical as the one they know. To the left [above], 'MM-7' helps to shed the coat sir.
Google translation from Spanish
LADY: BEHOLD YOUR NEW DOMESTIC
It's ugly, but does not answer to bad manners
THE Vienese Engineer Claus Scholz is about to solve the problem of domestic service, as alive as living standards improve. He has built and what will be the perfect servant and that currently called "Selector MM-7 '. This is a "robot" able to perform simple tasks, such as dusting, open the door, sweep, serving snacks and dry dishes. The mechanical maid is high. It measures 1.82 meters, but weighs little: is less than 50 kilos. Her skin is not pleasant, it is made of synthetic resin, and has feet too big, first to fit the number 49. Nor can it be said to offer a pleasant face. But evil does not answer, no armhole and, above all, does not require frequent wage increases or claims free time to go with her boyfriend to "cinema". For now, the new home is in the period that the housewives have devoted to updating the servants from the village. The engineer Scholz says his "chacha" soon be perfected and that before long enough that the lady says, "Prepare a meal for six people" so that the "robot" fulfill the order. For now, the acquisition of a maid is expensive electronics. If forecasts are met rows mass production and further desired, housewives can calm breathing, and tightness of the traditional "girls serve" will be corrected with the entry into the homes of these dolls, which provided "mnemonics files" receive instructions and develop relevant work .., without question. So be it.
Source: Blanco y Nero, Dec 1963
MM7 as he was in 2009.
In May of 2009 I travelled from Australia to Europe on a pilgrimage to see the old robots the had infuenced and aroused me in my youth. I went to Vienna to see Heinz Zemanek and his many Cybernetic tortoises and Maze solvers. These items are now housed in the Technical Museum of Vienna. Whilst there, I also saw Scholz's MM7, which was unexpected. DI Dr. Otmar Moritsch arranged for my behind the scenes visit. Here is the address of the Museum:
It looks as if Scholz had walking and stability problems early on with MM7. The knees were then "frozen" to prevent a bending action, and a metal frame added that also extended the length of the feet and added a steering castor at the rear.
David Buckley (on the right) joined me on my pilgrimage. Peter Schoen, Dr Otmar Moritsch's colleague who went out of his way to ensure our visit was a success, is on the left.
Close-up of the hands.
Detail of the head.
MM7 has no back as such. His exoskeleton body is made of reinforced fibreglass. MM7 was mains, not battery powered.
MM9 was built around 1973, so i'm lead to believe, but currently I have no image of it. MM6 was built in 1957-8.
Im Technischen Museum Wien ist noch bis 14. Juli 2013 die Ausstellung "Roboter. Maschine und Mensch?" zu sehen.
Christian Stadelmann, gemeinsam mit Bodo-Michael Baumunk Kurator der Ausstellung, stellt hier eines seiner Lieblingsexponate vor. Es handelt sich um den „Maschinenmenschen“ Numero 6 aus dem Jahr 1958. Erbaut wurde er von dem Wiener Kybernetiker Claus Scholz-Nauendorff in dessen Privatwohnung. Christian Stadelmann umreißt die Bedeutung dieses Roboters folgendermaßen:
„Das ehrgeizige Ziel, das Scholz-Nauendorff verfolgte, war es, sogenannte künstliche Intelligenz in Gestalt humanoider Roboter zu schaffen. Die Ergebnisse dieser Arbeit muten aus heutiger Sicht bizarr an, allzu simpel erscheinen die technischen Lösungen angesichts des Anspruchs, ein dienstbares Wesen zu schaffen. Aber die Öffentlichkeit zeigte vor allem an den Nachfolgemodellen MM7 und MM8 großes Interesse. Scholz Nauendorff präsentierte seine ‚Geschöpfe’ mit ernstem Stolz im Fernsehen und in Zeitungs- und Zeitschriftenartikeln. Dauerhafter Erfolg war diesen Forschungsaktivitäten nicht beschieden. MM7 kam ins Technische Museum Wien, MM8 ins Wiener Bezirksmuseum Landstraße. Sie haben dort einen eher skurrilen Status erlangt. MM6 ist überhaupt in Vergessenheit geraten.“
Und wie kam nun dieser MM6 in die Ausstellung des Technischen Museums? Das ist eine jener spannenden Geschichten, wie sie Ausstellungsmacher/innen auf der Suche nach interessanten Exponaten gerne erleben. Im Zuge der Recherchen zum Begleitmaterial zu den „Maschinenmenschen“ konnte Christian Stadelmann die Witwe des 1992 verstorbenen Kybernetikers, Friedericke Scholz-Nauendorff, ausfindig machen. Während eines Gesprächs über die „Maschinenmenschen“ ihres Mannes erwähnte sie dem Kurator gegenüber en passant, dass „einer von denen“ ja noch „hier herumsteht“. Für Christian Stadelmann war das ein ebenso aufregender wie berührender Augenblick:
„Auf meine unsichere Frage hin, was sie denn meine, führte mich die beim Gespräch anwesende Pflegerin von Frau Scholz-Nauendorff in einen Vorraum zur Küche, wo tatsächlich in einem Erker der mannshohe, über 50 Jahre alte Roboter stand. Wegen mechanischen Problemen hatte Scholz-Nauendorff seinerzeit die Entwicklung daran eingestellt und mit der Herstellung des Nachfolgemodells begonnen. Über meinen Wunsch, den MM6 ins Museum zu holen, zeigte sich die Betreuerin sehr erfreut, denn sie fürchtete sich jedes mal, wenn sie an ihm vorbeigehen musste, wie sie gestand.“
Für Christian Stadelmann ergab sich bei dieser Gelegenheit nicht nur ganz unerwartet ein tolles Ausstellungsobjekt, sondern auch eine direkte Verknüpfung zu den vielen Hoffnungen, die in Roboter gesteckt wurden und werden – wozu auch die Vorstellung zählt, dass Roboter dereinst in der Pflege von kranken Menschen eingesetzt werden könnten. Ein Gedanke, der in diesem Moment und in Gegenwart einer Pflegerin aus Fleisch und Blut einen ähnlich unheimlichen Beigeschmack besaß wie die äußere Gestalt des MM6.
English translation
ANSICHTSSACHE NO. 14: – ". Robot machine and man" "One of those" Christian Stadelmann about one of his favorite exhibits in the exhibition.
The Technical Museum in Vienna until 14 July 2013, the exhibition "Robots. Machine and man?" to see.
Christian Stadelmann, together with Bodo-Michael Baumunk curator of the exhibition, shares with us one of his favorite exhibits. These are the "human machine" Numero 6 from the year 1958. It was built by the Viennese cyberneticist Claus Scholz Nauendorff in his private apartment. Christian Stadelmann outlines the importance of this robot as follows:
"The ambitious goal of the Scholz-Nauendorff pursued, was to create so-called artificial intelligence in the form of humanoid robots. The results of this work seem bizarre at from today's perspective, the technical solutions seem overly simplistic, given the claim of creating a serviceable creature. But the public was mainly due to the subsequent models MM7 and MM8 interest. Scholz Nauendorff presented his 'creatures' with earnest pride on television and in newspaper and magazine articles. Long-term success was not granted these research activities. MM7 came to Vienna Technical Museum, MM8 into Vienna's Museum highway. They have acquired a rather bizarre state. MM6 is ever forgotten. "
And how does this MM6 came into the exhibition the Museum of Technology? This is one of those fascinating stories, as curator / inside in search of interesting exhibits like to experience. In the course of research on the supporting materials for the "machine-man" could make Christian Stadelmann's widow died in 1992 cyberneticist, Friedericke Scholz Nauendorff, locate. During a conversation using the "machine-man" of her husband, she mentioned the curator over en passant that "one of those" still "around here is." For the Christian Stadelmann was as exciting as touching moment:
"In my uncertain question what they mean for me the nurse present during the conversation of Mrs. Scholz Nauendorff led into an anteroom to the kitchen where actually stood in an alcove of the head-high, over 50 years old robot. Due to mechanical problems Scholz Nauendorff had once stopped developing it and started the production of a new model. About my wish to bring the MM6 to the museum, the supervisor was very pleased, because she was afraid every time she had to walk past him, as she confessed. "
For Christian Stadelmann was found on this occasion not only quite unexpectedly a great exhibit, but also a direct link to the many hopes that have been placed in robot and be – including the idea is one that robot one day used in the care of sick people could be. A thought that at this moment and in the presence of a nurse in the flesh had a similar sinister connotation as the exterior of the MM6.
As a result of the 2012-13 Robot exhibition held at Vienna Technical Museum, I now have images of MM6.
Scholz was exploring electro-hydraulic limbs in MM6, dating from 1957-58.
Above photo by Thomas Preiss.
Images by David Kotrbar.
MM8 normally resides at Wiener Bezirksmuseum Landstraße (Vienna's District Roads Museum).
MM8 doesn't walk (slide) as per MM7, but rolls around with rigid legs.
Picture rights: praktiker.at/Felix Wessely
Claus Scholz with MM8 in 1990.
Picture is from praktiker magazine 6/1990 which carries an excellent article on the story of Claus Scholz and has diagrams of MM6 to MM9.
Claus was a Professor. The only paper I could find is "System with automated exploring of problems for intelligent data processes by ergonomic dialog" published in 1982 under his fulll name of Claus Christian Scholz-Nauendorff.
Audio files by Scholz are located here and here. An English translation by a reader would be greatly appreciated and acknowledged.
Missing Articles: Mentioned here are 2 articles on Scholz that I'm having difficulty in locating.
One is supposedly in an 1964 edition of LIFE magazine. I've searched all of 1964 issues online and not found this article. It may be in an International edition of Life Magazine.
The other is in a 1970 German edition of Mickey Mouse (Micky Maus).
Any assistance in locating these articles would be much appreciated.
Fig. 38. Schematic of the charge, voltage conversion, the element changes tropism and chain contact device "turtle" "Tortilla".
Fig. 39. Schematic of extreme search direction of the "turtle" "Tortilla".
Fig. 40. Schematic of reaction "turtle" "Tortilla" with the whistle.
Information and images courtesy Waldemar Dekański from Poland (January 2010).
Hello Reuben!
I'm sending you Tortilla materials just received from Ukraine. It's part of a book by A. Yvahnenko "Technical Cybernetics". According to data the project and construction of the turtle was done by three engineers from Automatics Laboratory of Electrotechnical Institute in Kiev: T. Kravec, Y. Krementulo i E. Shukaylo. I presume it was built in 1958, the book describing the turtle was published in 1959. In the same year article was issued by J.Krementulo in "Automatika" magazine. I'm during intensive search for that article.
Cheers, Waldemar.
——————————————
Кибернетическая черепаха. Рассмотрим еще пример программной системы, где самоизменение программы подчиняется не одному, а нескольким требованиям. Таким примером может быть «черепаха» Вальтера [9], [59], [50]. «Черепаха» представляет собой автоматическую игрушку,
и
воспроизводящую все основные черты поведения живой черепахи. Конструктивно она выполнена в виде небольшой тележки на трех колесах, на которой установлены два сервомотора (ход вперед и поворот), электромагнитные
и
реле, электронная аппаратура и питающий аккумулятор.
Если аккумулятор хорошо заряжен, то «черепаха»
ведет себя как сытая и ищет темный угол в комнате. Если
аккумулятор разряжен, то «черепаха» ищет кормушку.
Такой «кормушкой» служит место для зарядки аккумулятора, освещенное сильной электрической лампой.«Черепаха» ищет свет и, подойдя к месту зарядки, стоит там пока не зарядится аккумуляторы. Затем снова уходит в более темное место комнаты.
Первые «черепахи» Вальтера (под названием «Элси» и «Элмер») реагировали на источник света только в зависимости от состояния своего «желудка» (аккумулятора).
443
В следующей разработке («черепаха» «Кора») автор осуществил еще добавочную реакцию на свист. При свисте «черепаха» замирает, т. е. некоторое время не движется. Если свист повторяется весьма часто, то «черепаха» перестает на него реагировать и продолжает либо искать «кормушку», либо уходит от нее.
Если «черепаха» наталкивается на препятствия, то программа ее действий изменяется (элемент самоизменения программы). Она делает ход назад, поворот, а затем только продолжает поиск «кормущки».
Правила действий (алгоритм) «черепахи» можно записать в виде табл. 5.
В табл. 5 сигналы расположены по силе их действия. Сигнал от контактного датчика имеет преимущество перед сигналом фотоэлемента, а сигнал от микрофона действует сильнее всех других сигналов.
124
Из таблицы следует, что главными программами являются: программа N9 1, обеспечивающая поиск источника света, и программа М 2, обеспечивающая более быстрое движение «черепахи» по направлению к источнику света или от него. Каждая из этих программ может иметь ряд вариантов (количество ходов и величина их не оговаривались выше). Из вариантов программы тот лучше, при котором :
а) «черепаха» быстрее находит наиболее яркий источник света;
6) найдя источник, возможно быстрее движется к немо.- (или от него).
Важно также, чтобы «черепаха» наиболее точно выполняла требования, указанные в таблице, и не теряла источника света из своего поля зрения, т. е. чтобы, перейдя к программе No 2, не возвращалась снова где-либо в пути к программе М. 1. Таким образом, «черепаха» имеет несколько показателей качества программы, кроме того, ее движение еще подчинено ряду дополнительных требований (ограничений).
Ниже мы рассмотрим более подробно схемы управления «черепахи», удовлетворяющие этим требованиям.
После «черепах» английского инж. Вальтера автоматические «черепахи» разрабатывали австрийский инж. Земанах, немецкий инж. Эйхер и др.
В СССР различные конструкции «черепах» разрабатывались в Институте автоматики и телемеханики АН СССР (инж. А. М. Петровский и Р. Б. Васильев), в Московском инженерно-физическом институте, в Институте автоматики Грузинской ССР и др. «Черепаха» «Тортилла», описываемая ниже, разработана в лаборатории автоматического регулирования Института электротехники АН УССР. Экспериментальная часть выполнена инженерами Т. Д. Кравцем, Ю. В. Крементуло и Е. И. Шукайло.
С точки зрения техники экстремального регулирование основная программа «черепахи» может быть решена двумя различными способами :
1) при помощи системы колебательного экстремального поиска наиболее яркого места горизонта, осуществляемого одним фотоэлементом («черепаха» «Тортилла-1 ») ;
2) при помощи неколебательной обратной связи, осуществляемой двумя фотоэлементами, направленными под
125
небольшим углом в деве соседние точки горизонта («черепаха» «Тортилла-2»).
В последнем случае мы располагаем всеми точками экстремальной характеристики одновременно и потому можно осуществить систему неколебательного установления экстремума (подробнее см. выше) .
Колебательная система благодаря наличию фильтра более помехоустойчива. Неколебательная система проще и надежнее.
Для краткости дадим описание только «черепахи» «Тортилла-1» (с колебательным поиском)1.
На рис. 38 изображена схема экстремального регулирования направления движения «черепахи» «Тортилла-1». В ней применена система шагового экстремального регулирования, рассмотренная в предыдущей главе.
Система экстремального поиска «черепахи» «Тортилла-1» действует следующим образом. Напряжение, вырабатываемое фотоэлементом ЦГ-4, усиливается при помощи электронного усилителя и поступает затем на контактные устройства шагового распределителя ШР, имеющего четыре поля. Цикл работы системы весьма прост. На первом контакте второго поля шаговый распределитель производит стирание предыдущей записи с первого электронного запоминающего устройства 3У1, а вторым контактом первого поля производится на нем новая (первая) запись напряжения. Третий контакт второго поля осуществляет стирание записи со второго запоминающего устройства 3У2, а третий контакт четвертого поля включает напряжение на сервомотор СМ1, который поворачивает фотоэлемент на шаг 7,5°. После этого четвертым контактом первого поля производится вторая запись усиленного напряжения фотоэлемента на 3У2, а пятым контактом третьего поля – сравнение напряжений первой и второй записи. Элемент логического действия ЭЛД включает сервомотор СМ1 в направлении, обеспечивающем движение (вращение) фотоэлемента к направлению экстремальной (наибольшей или наименьшей) освещенности. затем цикл операций повторяется сначала.
одиннадцатый и двенадцатый контакты четвертого поля (рис. 39) используются для: а) включения напряжения на
1 «Черепаха» «Тортилла-2» описана Ю. В. Крементуло в журнале «Автоматика», Х2 2, 1959.
126
сер вомотор СМ2 продольного перемещения «черепахи»; б) замыкания на короткое время цепи реле перемены тропизма 1; в) перехода от программы \5 3 к программе \5 1 или 2 в случае, если «черепаха» встретила препятствие (см. табл. 5); г) для подачи импульсов на схему реакции «черепахи» на звук (рис. 40). Измерительным элементом системы служит мост М с двумя стабиловольтами СГ-ЗС см. рис. 38,. При определенном напряжении (выбираемом Рис. 40. Схема реакции «черепахи» «Тортилла» на свисток. произвольно путем установки тех или иных сопротивлении моста) напряжение на выходе моста изменяет знак, что и приводит к переключению поляризованного реле перемены тропизма РП.
Поляризованные реле РП2 и РПз образуют элемент логического действия ЭЛД по схеме равнозначности.
зарядка аккумулятора производится через контактную шину КШ и релерегулятор РР. Напряжение постоянного тока аккумулятора при помощи вибропреобразователя ВП преобразуется в высокое напряжение переменного тока. Последнее выпрямляется и используется для питания
Под переменой тропизма «черепахи» понимается переход от поиска света к поиску темноты и наоборот, в зависимости от напряжения аккумулятора.
9 441 129
анодов 3У и усилителя фототоков. Датчик препятствий ДП при встрече с «черепахой» какого-либо препятствия срабатывает и при помощи реле РП,4 изменяет программу хода вперед на программу хода назад. В этом случае «черепаха» делает один шаг назад (на одиннадцатом контакте) и некоторое время двигается по направлению, перпендикулярному с направлением на источник света. Это достигается включением вместо основного вспомогательного
Рис. 41. Общий вид «черепахи» «Тортилла-2» .
фотоэлемента, направленного перпендикулярно оси «черепахи». Режим обхода препятствий кратковременный : как только подвижный контакт шагового распределителя дойдет снова до 12-ой ламели, то, как видно из схемы (рис. 39), основная программа «черепахи» восстанавливается.
Частота импульсов определяет собой скорость действий «черепахи». Скорость передвижения «черепахи» оказывается достаточной, если полный оборот распределителя происходит за б сек. В качестве генераторов импульсов можно использовать контактное устройтво, вращаемое отдельным двигателем.
Рассмотрим теперь действие цепи, осуществляющей реакцию «черепахи» на свисток (рис. 40). В качестве микрофона М использована пьезоэлектрическая телефонная трубка. Схема резонансного усилителя подобна схеме акустического управления радиоприемником, описанной в журнале
130
«Радио», М 4 за 1957 г. Реле Р1 на выходе схемы срабатывает под действием звука (свисток с частотой около
9000 гц) и останавливает оба сервомотора СМ1 и СМ2 «черепахи» (рис. 39).
Время остановки «черепахи» определяется параметрами нагрузки (1? и С) детектора. Если свистки повторяются редко, то конденсатор С успевает разряжаться, реле Р1 отпускает контакт и «черепаха» начинает снова двигаться. Если же свистки следуют часто, то напряжение на обмотке реле Р1 подымается выше некоторого предела, срабатывает реле Р2, шунтирует контакт реле Р1 и «черепаха» перестает реагировать на свистки. Блокировка реле Р2 снимается основным распределителем при прохождении через 12-ый контакт, если конденсатор к этому времени достаточно разрядится.
График типичного пути «черепахи» «Тортилла-1» к источнику света представляет собой ломаную линию. Общий вид «черепахи» «Тортилла» представлен на рис. 41.
Данные элементов «черепахи» «Тортилла» приведены па рис. 38-40.
Шаг поворота фотоэлемента составляет величину от 7,5 до 60° при частоте импульсов от 0,5 до 3 импульсов/сек. «Черепаха» реагирует на источник света (лампа накаливания мощностью 25 вт) на расстоянии до 3 м.
Некоторые дополнительные технические данные «черепахи» «Тортилла-1»
РП – поляризованное реле типа РП;
СМ – двигатели па 24 или вит. 27 в;
ШИ – шаговый искатель;
Тр – трансформатор, имеющий:
= 2 х 60 вит; д1 = 0,6 мм;
W2 = W3 = 3000 вит; д23 = 0,12 мм; В1В2 – выпрямители, собранные на ДГ-Ц24;
Б – аккумулятор типа 5 НКН-10: Напряжение тахогенератора 6 в.
—————
English Translation
Cybernetic tortoise. Consider another example of a software system, where self-transformation program obeys no one, but several requirements. An example might be "tortoise" Walter [9], [59], [50]. "Turtle" is an automatic toy
and
reproducing all the main features of the behavior of living turtles. Structurally, it has been implemented in the form of a small truck on three wheels, in which there are two servo-motor (move forward and turn), electromagnetic
and
relays, electronic equipment and power supply battery.
If the battery is well charged, the "turtle"
behaves as a well-fed and looking for a dark corner in the room. If
the battery is discharged, the "turtle" is looking for a manger.
Such a "trough" is a place to charge the battery, illuminated by a strong electric light. "Turtle" is looking for the light and going to a place charging stands there until you charge the batteries. Then again takes place in a dark room.
The first "turtle" Walter (called "Elsie" and "Elmer") reacted to the light source only depending on the state of his "stomach" (battery).
443
In the next development ( "turtle" "bark"), the author conducted more additional responses to the whistle. When whistling of "turtle" freezes, ie, some time not moving. If the whistle is repeated very often, the "turtle" ceases to react to it and continues to seek a "feeder" or away from it.
If the "turtle" is impeded, the program changed its course of action (element of self-transformation program). She makes a move back, turn, and then just continues to search for "kormuschki.
Terms of action (algorithm) "turtle" can be written in the form of tables. 5.
Table. 5 signals are located on the strength of their actions. The signal from the contact sensor has an advantage over the photocell signal and the signal from the microphone effect is stronger than all the other signals.
124
The table shows that the main programs are: Program N9 1, provides a search of the light source, and the program of M 2, which provides a more rapid movement of "turtle" in the direction of the light source or away from him. Each of these programs may have a number of options (number of moves and not subject to value them above). Of the options program that is better, in which:
a) "turtle" quickly finds the most brilliant source of light;
6) finding the source as quickly as possible moves to dumb .- (or him).
It is also important to "turtle" most closely meet the requirements listed in the table and not lose the light source from its field of view, ie that by going to the program No 2, did not return again, somewhere in the path of the program M. 1. Thus, the "turtle" has a quality program, in addition, its movement is still subject to a number of additional requirements (constraints).
Below we consider the more detailed management scheme "turtle", satisfying those requirements.
After the "turtle" the British engineer. Walter automatic "turtle" develop an Austrian engineer. Zeman, a German engineer. Eyher etc.
In the USSR, various constructions of "turtles" were developed at the Institute of Automation and robot USSR (Ing. A. Petrovsky, R. B. Vasiliev), at the Moscow Engineering Physics Institute, the Institute of Automation of the Georgian SSR, etc. "Turtle" Tortilla ", described below, was developed in the laboratory of automatic control of the Institute of Electrical Akad. The experimental part is made by engineers TD Kravtsov, V. Krementulo and EI Shukaylo.
From the standpoint of extreme technology management core program of "turtle" can be solved in two different ways:
1) with the help of vibrational find the most extreme places of the bright horizon of single photocell ( "turtle" "Tortilla-1");
2) using nonoscillatory feedback undertaken by the two photocells directed at
125
slight angle to the horizon Maiden neighbor ( "turtle" "Tortilla-2").
In the latter case, we have all the points of extreme characteristics simultaneously, and therefore can be carried out to establish a system of nonoscillatory extremum (see above).
Oscillatory system thanks to the filter more robust. Nonoscillatory system is simpler and more reliable.
For brevity, only give a description of "turtle" "Tortilla-1" (with vibrational search) 1.
Fig. 38 is a diagram of extremal control the direction of "turtle" "Tortilla-1". It used a system of extremal control step considered in the previous chapter.
System of extreme search for "turtle" "Tortilla-1 operates as follows. Voltage, provided by a photocell CG-4, augmented by an electronic amplifier and then fed to the contact device stepper distributor WAF, which has four fields. The cycle of the system is very simple. At the first contact of the second field stepper valve makes erasing the previous record with the first electronic storage device 3U1, and the second contact of the first field is made on it new (first) record voltage. Third contact, the second field carries erasing records from the second storage device 3U2, and the third contact, the fourth field includes the voltage on the servo motor CM1, which turns the photocell step 7,5 °. After this, the fourth pin of the first field is the second record amplified voltage to the photocell 3U2, and the fifth contact, a third of the field – a comparison of the stress of the first and second record. Element of the logical steps ELD includes servo CM1 in the direction of securing the movement (rotation) of the photocell to the direction of the extreme (highest or lowest) illumination. then the cycle of operations is repeated again.
eleventh and twelfth contacts of the fourth field (Fig. 39) are used to: a) the inclusion of voltage
1 "Turtle" "Tortilla-2" described YV Krementulo in the journal "Automation", A2 2, 1959.
126
Ser vomotor SM2 longitudinal movement of "turtle", and b) circuit for a short time relay circuit changes tropism; 1) the transition from the program \ 5 3 to the program \ 5 1 or 2 if the "Tortoise" obstacles encountered (see Table. 5) d) to supply pulses to the reaction scheme "turtle" to the sound (Fig. 40). The measuring element of the system is a bridge with two M stabilivolt SG-AP, see Fig. 38. At a certain voltage (selectable Fig. 40. Scheme of the reaction of "turtle" "Tortilla" on the whistle. Arbitrarily by setting the resistance of some of the bridge) the bridge output voltage changes sign, which leads to a shift of the polarized relay RP tropism changes.
Polarized relay IS2 and RPZ constitute an element of logical steps ELD scheme equivalence.
Charging the battery is made through the contact bus SH and releregulyator PP. Voltage DC battery with vibrator MP is converted to high voltage alternating current. Last rectified and used to power
Under the change of tropism "turtle" refers to the transition from search to search the world of darkness and vice versa, depending on battery voltage.
9 441 129
anodes 3U and amplifier photocurrents. Sensor obstacles DC at a meeting with the "turtle" is no impediment, and is triggered by relay RP, 4 modifies the program moves ahead on the program of the back. In this case, "turtle" makes one step back (at the eleventh contact) and for some time moving in a direction perpendicular to the direction of the light sources. This is achieved by inserting instead of the main support
Fig. 41. General view of the "turtle" "Tortilla-2".
photocell directed perpendicular to the axis of "turtle". Mode to avoid obstructions brief: as soon as the movable contact stepper distributor comes back to the 12th slats, then, as seen from the scheme (Fig. 39), the main program "Turtles" is restored.
The frequency of pulses determines the speed of action "Turtles". Speed of movement "turtle" is sufficient, if the total turnover of the distributor is used for the second. As pulse generators can use the contact ustroytvo waved in a separate engine.
We now consider the effect of the chain, carrying out the reaction of "turtle" on the whistle (Fig. 40). As the microphone M used piezoelectric handset. Scheme of the resonant amplifier circuit is similar to the acoustic / radio, described in the journal
130
"Radio", No. 4 for 1957 Relay R1 at the circuit output is triggered under the effect of sound (a whistle with a frequency of about
9000 Hz) and stops the servomotor both CM1 and SM2 "turtle" (Fig. 39).
Time stop "turtle" is determined by the parameters of load (1? And C) detector. If the whistles are rarely repeated, the capacitor C has time to be discharged, the relay R1 releases the contact, and "turtle" again begins to move. If, however, often followed by the whistles, the voltage across the relay coil R1 rises above a certain limit, relay P2, shunts the relay contact P1 and "turtle" ceases to respond to whistles. Blocking relay P2 is removed the main distributor in passing through the 12th contact, if the capacitor at that time sufficiently discharged.
Schedule a typical path of "turtle" "Tortilla-1" to the light source is a broken line. General view of the "turtle" "Tortilla" is presented in Fig. 41.
These elements of the "turtle" "Tortilla" shown in Fig. 38-40.
Step turning the photocell is a quantity from 7.5 to 60 ° at a frequency of pulses from 0,5 to 3 pulses / sec. "Turtle" responds to the light source (incandescent lamp of 25 W) at a distance of 3 m.
Some additional technical data "turtle" "Tortilla-1"
RP – polarized relay type RP;
SM – engines pas 24 or vitamin. 27 in;
SHI – step seeker;
Tr – transformer with:
= 2 x 60-vit; D1 = 0.6 mm;
W2 = W3 = 3000 vitamin; d23 = 0.12 mm; V1V2 – rectifiers, gathered at the DW-TS24;
B – Battery type 5 ICH-10: Tacho Voltage 6.
-x640.jpg)
