technology and autonomous mechanisms in the mediterranean: from ancient greece to byzantium...

110 IEEE CONTROL SYSTEMS MAGAZINE » DECEMBER 2014

James Watt’s flyball governor (1769) is one of the earliest feedback con-trol devices of the modern era.

This device was used to regulate the

rotational speed in steam locomo-tives and became very useful during Europe’s Industrial Revolution in the 1800s. As demands on the device grew, the flyball governor was required to work in new operating regions where it exhibited undesirable oscillations for reasons unexplained at that time. J.C. Maxwell (1868) used differen-tial equations to model and explain this phenomenon, which marked the beginning of the introduction of math-ematical rigor into the study of feed-

back control devices, together with the works of Routh (1877) and Lyapunov (1890) [1]–[9].

However, the discipline of feed-back control has a very long history, dating back more than 20 centuries. Automatic feedback control devices appeared in the work of Ktesibius (or Ctesibius, also Ktesivios) of Alex-andria (circa 270 BC) as reported by Vitruvius (circa 25 AD), Heron (or Hero) of Alexandria (circa 60 AD), and Philon of Byzantium (circa 230 BC).

Technology and Autonomous Mechanisms in the Mediterranean: From Ancient Greece to Byzantium

KIMON P. VALAVANIS, GEORGE J. VAChTSEVANOS, and PANOS J. ANTSAKLIS

This contribution describes the au-

tomation and control technology of

ancient times.

Digital Object Identifier 10.1109/MCS.2014.2359589

Date of publication: 13 November 2014

written on robust design methods, but these typically address the robustness of system stability to model variations rather than performance concepts desired by the practicing engineer. Indeed, without some new concepts, the theoretical solution of some prob-lems appears impossible. For example, is it possible to find the bounds on two parameters of a transfer function so that the overshoot in its step response is never greater than 15% without solv-ing for the step response explicitly? This problem is easily “solved” using Monte Carlo methods in simulation. Fundamentally, the lack of an analyti-cal solution seems to be due to limita-tions on what can be learned from the s-domain about time-domain perfor-mance. Current simulation facilities are excellent, and I am sure they will remain the major design tool, together with basic control engineering theoret-ical knowledge, of the practicing engi-neer for many years to come.

Without doubt, no one could have predicted the advances of the last 60 years, so it is foolish to do so for the future, but what is certain is that, with more than 90% of qualified scientists

and engineers who have ever lived alive today, further technological progress is undoubtedly forthcoming. What these developments will be, however, is not clear, Whether the developments are in energy, new processes, new materials, or transportation, what is certain is that control technology— “the hidden tech-nology,” to quote Karl Åström—will have a role to play.

REFERENCES[1] T. E. Broadbent, Electrical Engineering at Man-chester University. Manchester, U.K.: Univ. Man-chester, 1998.[2] B. Chance, V. Hughes, E. F. MacNichol, D. Sayre, and F. C. Williams, Waveforms (MIT Ra-diation Lab Series, vol. 19). New York: McGraw-Hill, 1947.[3] B. Chance, F. Hulsizer, E. F. MacNichol, and F. C. Williams, Electronic Time Measurements (MIT Radiation Lab Series, vol. 20). New York: McGraw-Hill, 1947.[4] S. Bennett, “F. C. Williams: His contribution to the development of automatic control,” IEE Electron. Power, vol. 25, no. 11, pp. 800–802, Nov./Dec. 1979.[5] H. M. James, N. B. Nichols, and R. S. Phillips, Theory of Servomechanisms (MIT Radiation Lab Series, vol. 25). New York: McGraw-Hill, 1947.[6] J. C. West, Servomechanisms. London, U.K.: English Univ. Press, 1953.[7] J. C. West, Analytical Techniques for Nonlinear Control Systems. London, U.K.: English Univ. Press, 1960.

[8] D. P. Atherton, “Non-linear systems with mul-tiple inputs,” Ph.D. dissertation, Dept. Elec. Eng., Univ. Manchester, Manchester, U.K., 1962. [9] D. P. Atherton, “Some reflections on analogue simulation and control engineering,” Meas. Con-trol, vol. 37, no. 10, pp. 300–306, 2004.[10] D. P. Atherton, “Control engineering and the analog computer,” IEEE Control Syst. Mag., vol. 25, no. 3, pp. 63–67, 2005.[11] F. R. J. Spearman, J. J. Gair, A. V. Hemingway, and R. W. Hynes, “TRIDAC, a large analogue computing machine,” Proc. IEE, vol. 103, pt. B, pp. 375–395, Oct. 1955.[12] J. H. Westcott, “Automatic control. The I.F.A.C. congress, Moscow, 1960,” J. Inst. Electr. Eng., vol. 6, no. 72, pp. 732–733, Dec. 1960.[13] D. P. Atherton, Nonlinear Control Engineering: Describing Function Analysis and Design. London, U.K.: Van Nostrand Reinhold, 1975.[14] D. P. Atherton. (2012). An Introduction to Nonlin-earity in Control Systems. Denmark: Bookboon Pub-lications. [Online]. Available: www.bookboon.com[15] D. P. Atherton, “Conditions for periodicity in control systems containing several relays,” presented at the Proc. 3rd IFAC World Congr., London, Paper 28E, June 1966. [16] Y. Tsypkin, Theorie der Relais System der Au-tomatischen Regelung. Munich, Germany: Olden-bourg-Verlag, 1958.[17] J. K. C. Chung and D. P. Atherton, “The de-termination of periodic modes in relay systems using the state space approach,” Int. J. Control, vol. 4, no. 2, pp. 105–126, 1966.[18] F. C. Williams and J. C. West, “The position syn-chronization of a rotating drum,” Proc. IEE—Part II: Power Eng., vol. 98, no. 61, pp. 29–34, Feb. 1951.[19] Hard disk drive. (2014, 29 Sept.). Wikipedia. [Online]. Available: http://en.wikipedia.org/wiki/Hard_disk_drive

1066-033X/14©2014iEEE

DECEMBER 2014 « IEEE CONTROL SYSTEMS MAGAZINE 111

The first feedback control device on historical record is the water clock of Ktesibius [1], [2], although Ktesibius’s contemporary Archimedes also had a water clock driven by a feedback mechanism.

Throughout history, humans have dreamed of devices that would remove the burden of heavy or repeated labor, offer protection from danger, and pro-vide amusement or entertainment. The idea of automatic devices, called automata ( )ayxqnaxal by the ancient Greeks, is well documented in Greek mythology and poetry. For example, myths refer to the god Hephaestus as the owner and builder of automated mechanisms, called golden women, who helped him in his machine shop and who were able to talk. Hephaestus also built a series of tables with three legs and wheels. Homer’s Iliad refers to these mechanical devices as the gold tri-pods of Hephaestus, or the golden slaves, that helped him to walk but also served wine and food to the gods of Mount Olympus [10]. Homer’s Odyssey also mentions Phoenician ships that moved by themselves as if pushed by an inter-nal force. Greek mythology refers to the giant Talus (T oam wl or ),Tam~wl made of brass, built by Hephaestus, and offered as a wedding gift to Europa when she married Zeus.

In some legends, Hephaestus was assisted by Daedalus ( ),oakdam wD l who later designed wings for himself and his son Icarus ( ),olat wIl in an attempt to fly over the Aegean Sea [10]. The leg-end goes that Zeus gave Talus to King Minos of Crete, who used the giant to convey his messages and laws all over the island, running back and forth three times a day, but also to protect the island against intruders by throw-ing rocks and molten lava at their ships. Thus, Talus had all the desired charac-teristics of a humanoid robot, includ-ing strength, service to humans, and the ability to multitask and perform repetitious and heavy tasks endlessly and without fatigue. After the mythical Talus, it was not until the 16th century when the Italian Hieronymus Fabrikius (1537–1619 AD), a famous anatomist and

embryologist of the time, attempted to build the first mechanical man using surgical tools [10].

The roots of automation, automated mechanisms, and feedback control in the real world go back more than 2500 years, and are centered in the ancient Greek-speaking Mediterranean re-gion. The term “machine” nh|aohl^ h with its modern meaning was used by Aeschylus (fifth century BC) to de-scribe the theater mechanism built to bring heroes in ancient Greek tragedies onto the stage (the Latin term for this mechanism, deus ex machina, was intro-duced by Vitruvius). The mechanism consisted of beams, wheels, and ropes that could raise weights up to one ton and move back and forth violently to depict flight, as the play demanded [11, pp. 473–478]. However, the term “system” ( )vyvxhnal had different meanings in geometry, numbers, phi-losophy, music, and medicine [12], but the basic notion that the whole is made from parts and that the whole has properties beyond the individual parts was common to all domains. Although some understanding of the “system” notion is due to Plato, Aristotle, and Hippocrates, its origin is credited to the pre-Socratic philosophers and their effort to comprehend the unknown physical world. The first explicit defi-nition of system is attributed to the Lakonian and Pythagorean Kallicra-tides in his work “On the Happiness of Family” ( )O E oftk kl~o ydakn okawP l l l where he explains what is a system in terms of three examples: the sys-tem of dance in the singing societies; the system of the crew on a ship; the system of the family, where the people are next of kin: “Any system consists of contrary and dissimilar elements, which unite under one optimum and return to the common purpose.” Kal-licratides’ definition indicates that the system could be considered not as a simple open system, but as a closed-loop control system consisting of op-posite parts, united to the optimum (controller aiming at optimization) and returning to the common target (embodying feedback).

Many ancient autonomous devices were built to amaze and entertain roy-alty and to demonstrate the skill of the designers of the devices. Typically, these devices used a clockwork mecha-nism that worked in open loop, without feedback, and, once started, followed a prescribed set of steps in sequence. In contrast, this article focuses on ancient mechanisms that employed automatic feedback. Covering roughly from 600 BC to 600 AD, this historical perspec-tive provides an overview of many ancient mechanisms. It is not the pur-pose to provide specific details on the construction or operation of any of the mechanisms. The interested reader is referred to the many cited references. While this article covers the Mediter-ranean region, it is hoped that other authors will be inspired to describe the contributions of other civilizations in other parts of the world.

Articles on feedback mechanisms often begin with the question of what constitutes a feedback mechanism. This article follows the definitions provided in [1] and [2]. The water clock with feedback, associated with Ktesibius and Archimedes, is consid-ered to be the first example of a feed-back mechanism in historical records. While there may have been mecha-nisms that used feedback before the water clock, there has not yet been conclusive proof of any. There is an additional issue that adds to the confu-sion about what constitutes a feedback mechanism. Many systems incorpo-rate internal feedback, that is, feedback built into the system itself. For exam-ple, the shape of a boat’s hull is such that the boat returns to the upright position when disturbed. The tail of a weather vane provides internal feed-back to align the weather vane with the wind direction. What constitutes a feedback device is influenced by the choice of what is internal or external to the system and whether the feed-back is considered to be an internal or external mechanism. While this may be an interesting question, devices employing internal feedback will not be considered here.


UNCOVERING hISTORICAL INFORMATIONThere are several sources of knowl-edge about the science and technology of the ancient world. Archaeology is the study of material culture, that is, artifacts. For instance, excavations of ancient workshops and mines have recovered tools and material that shed light on ancient technological pro-cesses. Wall paintings reveal further information. Many ancient texts that have technological content have sur-vived; for example, Mesopotamian texts from the second millennium BC contain recipes for copper alloys and glass. Another source of information is ethnoarchaeology ( ),o o oio at|ak m ckaf l a which studies contemporary commu-nities that live in ways and use tech-nology similar to that of antiquity, like pottery making. Experimental archae-ology studies ancient technology by creating replicas, for example, ships, siege weapons, or other mechanisms. Archaeometry uses the tools of mod-ern physical science to study ancient artifacts, for instance, to pinpoint their time period or to reconstruct ancient trading networks [13].

There are several difficulties in the study of ancient mechanisms. Many ancient manuscripts have been lost. Some manuscripts are known only from references made to them in other manuscripts. The existence of mechanisms and descriptions of their operation sometimes has to be pieced together from this fragmentary docu-mentation. Incomplete evidence may sometimes lead researchers, particu-larly amateur historians, to unwar-ranted conclusions. The ancient world, like today, had its own science fictions. Just because the functionality of a device was described in broad terms does not necessarily mean that the device was ever constructed and oper-ated. This manuscript clearly differen-tiates between sources that describe a futuristic vision and well-accepted facts with plenty of references. Pictures of replicas are included where possible.

In some instances, archeologists uncover pieces of an actual mecha-

nism. One such celebrated case is the discovery in a shipwreck of the Anti-kythira mechanism in 1901, currently exhibited at the National Museum in Athens, Greece. It took experts a long time to fully understand the purpose and function of the Antikythira mecha-nism, which has 30 bronze gears; the device was apparently used to calcu-late astronomical positions. So even in the instances when parts of a mecha-nism exist, it may not be straightfor-ward to uncover its function.

The next section outlines the science and technology of the Mediterranean region. Then, details of achievements during each period are described, fol-lowed by accomplishments in auto-matic control and the transition from the ancient Greek world to the Greco-Roman era and finally to Byzantium.

hISTORY OF SCIENCE ANd TEChNOLOGYThere was an initial phase of imported influences in the development of ancient technology that reached the Greek states from the East (Persia, Babylon, and Mesopotamia) and prac-ticed by the Greeks up until the sixth century BC. It was at the time of Thales of Miletus (circa 585 BC) when a very significant change occurred. A new, and exclusively Greek, activity called “science” became more important than imported technology. In subse-

quent centuries, technology became more productive, and technological innovations surfaced through syner-gies between science and technology. New measurement devices, geometri-cal concepts, engineering, mechanics, and musical instruments influenced important application domains such as metallurgy, construction projects, architecture, military tactics, and machines [14]–[16].

One of the first inventions of the fifth century BC, which is not widely known but was very important, is the starting mechanism in ancient stadiums. The hysplex ( )yvrmhpl was designed to prevent untimely starts in races (see Figure 1). Quoting from [17, pp. 465–472]:

The system consisted of two hori-zontal ropes stretched in front of the waist and the knees of the runners. The ends of the ropes were tied to the peaks of verti-cal wooden posts implanted in mechanisms near the two ends of the starting threshold. The mech-anisms were controlled by the starter, who, standing at the back of the runners, could let, at the ap-propriate moment, the ropes to fall down, thus, permitting all the runners to start simultaneously.

A hysplex was reconstructed at the University of California, Berkeley, and successfully tested on July 1, 1996 at an international track and field meeting called the Nemean Games. The hys-plex worked correctly 52 consecutive times [17].

An even older mechanism, chron-icled around 580 BC, is an oil press operated by the weight of stones. Such an oil press was found by Ferdinand André Fouqué in 1879 on Therassia, a small island opposite Thera (a.k.a. San-torini). The only existing evidence of this ancient Greek design is on exhibit at the British Museum in London.

Ancient Greek science is divided into four major periods in terms of technical accomplishments and break-throughs [14]–[16]:1) The pre-Socratic period from 600

BC to about 400 BC.

Figure 1 A model of one of the two parts of the hysplex [17], [21]. The hysplex ( )yvrmhpl was invented in the fifth century BC. The com-plete design is illustrated in a Panatheneon amphora dating back to 344 BC. The photo-graph is courtesy of the Thessaloniki Science Center and Technology Museum.


2) The fourth century BC (400–300 BC), which was the time of Plato, Aristotle, and the Epicurean and Stoic philosophers.

3) The Hellenistic period (300–100 BC), which was the time of Euclid, Archimedes, and Apollonius. The Hellenistic period is the golden era of ancient science and tech-nology that followed the death of Alexander the Great (323 BC), in which Alexander’s successors ruled a large territory. During that period, more direct interac-tion took place between Greek and other civilizations.

4) From 100 BC to 600 AD, where Greek science was influenced by spiritual currents associated with the rise of Christianity. During this period, Greek science was passed to the Arabs and the Latin West. Science was epitomized, reorga-nized, and subjected to extensive commentaries.Mathematics as a discipline was

t ightly coupled to technological achievements. Mathematics was not only a theoretical discipline but also a practical tool in a wide range of applica-tions and as the means to engineering design. It is no coincidence that Thales, Pythagoras, Euclid, Archimedes, and Apollonius were mathematicians first. Archimedes utilized mathematics in a completely innovative fashion for the treatment of physical problems. Fur-thermore, the tradition of formal me-chanical treatises, including theoretical and applied mechanics, as distinct do-mains from the almost completely theo-retical mechanics of Archimedes was continued into the third century BC by Ktesibius, whose work, although lost, is known through the references of Vitru-vius and Philon of Byzantium. The most comprehensive of the extant mechanical treatises is that of Heron of Alexandria, a recognized mathematician and an ex-cellent mechanician.

ANCIENT GREEK SCIENCE ANd TEChNOLOGYThere were three main institutions in the ancient Greek-speaking world

that contributed to science and tech-nology: Plato’s Academy, Aristotle’s Lyceum, and The Library of Alexan-dria. Plato’s Academy and Aristotle’s Lyceum focused more on science; the institutions were very influential, operated as schools or universities, charged tuition, and were financially self-sustained. The Library of Alexan-dria (and the museum associated with it) was established around 300 BC and exerted long-lasting influence over many centuries. The Library of Alex-andria operated more like a research laboratory than a university and was financially dependent on the rulers of Egypt, the Ptolemies [14]–[16], [18].

Among the earliest contributors to ancient Greek science and technol-ogy were the members of the Miletian School, particularly Thales of Miletus. The main contributions of the Miletian

School were the distinction between the natural and the supernatural and the introduction of rational criticism and debate. The achievements of the period between Thales and Aristotle were advances in factual knowledge (such as anatomy and zoology), prob-lem formulation (Aristotle’s biological problems of reproduction and hered-ity), the application of mathematics to the understanding of natural phenom-ena (Pythagoreans and Plato), and the notion of undertaking empirical research (Hippocrates and Aristotle).

Thales is thought to be the first per-son to study pure geometric objects, such as circles, lines, and triangles. He was the first who considered the angle to be a separate, distinct, and autonomous mathematical entity, and he established the angle as the fourth geometric element, complementing the other three: length, surface, and volume. Thales never referred to cal-culations and results obtained for a specific example or case study, as done by the Egyptians and the Babylonians, but rather, his works were very gen-eral. Thales is considered to be the first theoretician in the field of mathe-matics with theorem-proving abilities.

In regard to mechanisms, techno-logical innovation was either related to the invention of machines that relied on external sources of power, such as animal or wind power, or automatic devices, such as the inventions of Heron of Alexandria described below.

The first major breakthrough con-tribution to autonomous mechanisms occurred during the era of Pythago-ras, who was Thales’ student for a few years. The contribution is attributed to the Pythagoreans and, in particular, to Archytas from the city of Tarantas (a.k.a. Tarentum) in southern Italy, known as Archytas the Tarantine and sometimes as the Leonardo da Vinci of the ancient world. Archytas contrib-uted to number theory but was also the first engineer. By applying a series of geometric notions and observations to the study of structures, links, and joints, he created the field of mechanics. Archytas not only drew mechanisms,

(a)

(b)

Figure 2 Catapults of the type used by Philip and Alexander the Great, which were used to gain advantage in battles against their en-emies. The catapults were perfected using mathematical formulas from Alexandrian me-chanics that related the size of the catapults’ components to the weight of the projectiles or the size of the bolts that they were about to shoot [11, pp. 545–548], [21]. The photo-graphs are courtesy of the Thessaloniki Sci-ence Center and Technology Museum.


he also built them. In 425 BC he is reported to have created the first unmanned aerial vehicle by build-ing a mechanical bird that could fly by moving its wings, getting energy from a mechanism in its stomach. This mechanical bird is reported to have been able to fly about 200 m!

The introduction of catapults changed the art of war. The first cata-pults appeared in Syracuse, dur-ing the time of the tyrant Dionysius the Elder (a.k.a. Dionysius of Syracuse, 430–367 BC), who created the first sci-entific research center [10]. The widest and most well-known use of catapults occurred during the reigns of Philip of Macedonia and his son Alexander the Great, which were perfected to gain advantage in battles against their ene-mies. Mathematical formulas of Alexan-drian mechanics related the size of the

components of a catapult to the weight of the projectiles or the size of the bolts that were about to be shot [11, pp. 545–548]. Figure 2 illustrates two types of catapult that changed martial tactics and dominated the battlefield for many centuries. Figure 3 depicts reproduced

designs of other types of catapults. Sim-ilar mechanisms are shown in Figure 4.

Several inventions are attributed to Archimedes [11], [19], [20]. Figure 5 shows Leonardo da Vinci’s drawing for the “architronito,” which was a steam gun invented by Archimedes, and a modern reproduction [21]. Figure 6 illus-trates the endless screw. Other inven-tions by Archimedes include cranes and solar mirrors. The steam gun created the foundation for the theory of hydro-dynamics. Information about the steam gun is due to the Italian Petrarca (1304–1374 AD), but it was Leonardo da Vinci who explained its operation and stated that it could throw a bullet weighing one talanton ,oxamaox ol^ “heavy coin”) about 1100 m.

Another invention attributed to Archimedes is the water clock, shown in Figure 7, which stood more than 4-m tall (although some research-ers claim that the water clock was invented by Ktesibius, Heron, Philon, or others). Three sources—Vitruvius, Pappos the Alexandrian, and an Ara-bic document [20]—provide evidence

(a)

(b)

Figure 3 Catapult models reproduced by N. Orfanoudakis [21]: (a) a catapult used to throw Greek fire ( ),yctqo rytl an in-flammatory mixture used in ancient times) and (b) the theoretical design of a porta-ble mechanism to throw Greek fire, com-posed of pipes coupled with a tormentum serving as a power source—its range was about 22 m. The photographs are cour-tesy of the Thessaloniki Science Center and Technology Museum.

(a)

(d)

(b)

(c)

Figure 4 Models of ancient weapons reproduced by N. Orfanoudakis [21]: (a) a mechanism composed of a catapult and Heron’s pump to throw Greek fire at a distance of about 30 m, (b) a repetitive multiarrow catapult with range of about 100 m, (c) another catapult used to throw Greek fire, and (d) a ballista that can throw two arrows simultaneously. The photographs are courtesy of the Thessaloniki Science Center and Technology Museum.

(a)

(b)

Figure 5 (a) Leonardo da Vinci’s drawing for the “architronito,” which is a steam gun invent-ed by Archimedes and (b) a reproduction by J.G. Sakas in 1981. The images are courtesy of i.C. Lazos and the Thessaloniki Science Center and Technology Museum [20], [21].


that the inventor was Archimedes. It is important to state that it looks like there was progress in this area before Ktesibius invented his water clock (see below) without knowing exactly when it started. But since Archi-medes (287–212 BC) and Ktesibius (285–222 BC) lived almost at the same time, it may be that either Ktesibius improved Archimedes’ clock, or the two invented the water clock indepen-dently of each other.

The School of AlexandriaDuring the golden era of the School of Alexandria [18], Greek thought evolved toward what Newton called natural philosophy. This turn of Greek thought culminated in the search for observa-tion, measurements, design, and con-struction. The School of Alexandria is directly linked with three Greek engi-neers of the Hellenistic world: Ktesib-ius, Philon of Byzantium, and Heron of Alexandria. Ktesibius and Philon lived in the second century BC, while Heron

lived in the first century AD. Their con-tributions strongly influenced Western science and technology.

Ktesibius (circa 270 BC) was a Greek physicist, inventor, and the first great figure of the ancient engineering tradition of the School of Alexandria. The discovery of the elasticity of air is attributed to him, as is the invention of several devices using compressed air, including force pumps and an air-powered catapult. His most famous invention, however, was an improve-ment of the clepsydra, or water clock, in which water dripping at a constant rate raised a float that held a pointer to mark time (the passage of hours). Another notable invention was a hydraulis ( )ydtaymkwl or water organ, in which air was forced through the organ pipes by the weight of water rather than by falling lead weights [16],

[22]. Unfortunately, Ktesibius’s writ-ings have not survived, and his inven-tions are known only from references by Vitruvius [22] and Heron of Alex-andria. Figures 8–10 illustrate inven-tions made by Ktesibius. Ktesibius’s water clock, shown in Figure 9, incor-porated a feedback mechanism that used a floating device as both a sensor and actuator that kept the water level approximately constant and ensured a constant flow of water and, conse-quently, accurate time keeping [1], [2]. This design was so successful that it was still used in Baghdad in 1200 AD when the Mongols conquered the city!

Philon of Byzantium (circa 230 BC) authored several books (many of which have survived) on levers, water clocks, pneumatic mechanisms, and military machines, among others. His major contribution relates to the subject of spur gears. The transfer of mechanical power and the multiplying or dividing effect of circular-motion machines are based on such gears. Philon’s treatise Pneumatica addressed key physical properties of liquids and gases sup-ported by a plethora of experiments. He also advanced knowledge of met-allurgy. Philon explained thoroughly Ktesibius’s contention regarding the replacement of animal fibers with thermally treated metal leaves. In essence, Philon advanced further the Alexandrian School’s developments in

(a)

(b)

0 1m

Figure 6 (a) A representation of the endless screw [20], [28] based on archaeological find-ings in Centenillo, Spain. (b) A reproduction of the endless screw that was used in the mines in Sotiel Coronada, Spain. The reproduction was completed in 1930 by a research team of the School of Archaeology, University of Liv-erpool [28]. The images are courtesy of i.C. Lazos.

Figure 7 A drawing of the water clock of Ar-chimedes [20], a complex mechanism with movements powered by water. (a) The fa-çade, (b) the cross section, and the (c) back view. The water clock stood more than 4-m tall and is attributed to Archimedes by Arabic documents and other sources of antiquity, al-though some historians credit the water clock to Heron or Philon. The figure shows repro-ductions by the German scientists Wiede-mann and Hauser. The images are courtesy of i.C. Lazos.

Figure 8 The water musical instrument of Ktesibius and rare terracotta model [11], [21]. This invention by the founder of the School of Alexandria was based on air forced through the organ pipes via the weight of water rather than by falling weights. Ktesibius’s writings have not survived, and his inventions are known only through references by Vitruvius and Heron of Alexandria. The images are courtesy of the Thessaloniki Science Center and Technology Museum.


automatic mechanisms started by Ktesibius. The Arabic translation of Philon’s Pneumatica is a compendium of such automatic devices based on properties of air, liquids, fire, floats, and gears. These devices offered a rich foundation for a new technology: auto-mation. Although Philon did not com-plete his work, he laid the groundwork for the European technological devel-opments of the next 1500 years [16].

Eight hundred years of tradition in ancient Greek science and technol-ogy culminated with the enormous work of Heron, who was a scientist of great breadth. Heron of Alexandria (circa 60 AD) taught at the Museum of Alexandria and presented in his books Automata (a.k.a. Automatopoietica) and Pneumatica, a summation of what was known about automation technolo-gies. Among his surviving writings, Automata lays the foundation of mod-ern automatic control, and it is the old-est known document that describes mechanical automatic systems capable of performing programmed move-ments [16], [23]. Automata deals with automatic machines that operate and move without human intervention by capitalizing on the properties of steam, liquid, and gaseous matter using com-plex mechanical systems and ingenious programming of movements. Every mechanical device was broken down to its constituent components, which, in turn, were explained in terms of levers given a basic geometric interpretation. Although the objective of Heron’s book

was applications, he used a systematic methodology to explain the operation of mechanisms.

Heron’s second book, Pneumatica, introduced steam as a major driving force. Heron did not invent the first steam-powered device but presented the basic notions, such as the lifting and driving force of steam as well as the properties of steam under compression and expansion. These two books are the oldest preserved in their original Greek manuscripts that describe applications of pneumatic and hydraulic control sys-tems comprising a systematic collection of machines dating from before and during his lifetime. These machines moved by using water, air, or steam pressure; produced sounds; and were intended to decorate public places and to amaze or serve the practical needs of the spectators.

Although Heron in his books describes a menagerie of mechani-cal devices or “toys,” such as sing-ing birds, puppets, coin-operated machines, a fire engine, and a water organ, his most famous description is of the steam-powered engine, of special significance to the science of automation because he presented an automatic means necessary to control the steam supply. The device consists of a sphere mounted on a boiler and supported by an axial shaft with two

canted nozzles that produce a rotary motion, as shown in Figure 11.

Among Heron’s inventions are “automatic theaters,” capable of con-ducting an entire theatrical perfor-mance with automatic portals, chang-ing automatically the stage props, and moving puppets that performed a myth. For these mechanical automata, Heron wrote, “A prop is placed on a pylon depicting a theatrical scene with portals that open and close automati-cally. Therefore, it allows several scenes of a play to be enacted in sequence and the puppets are able to move, act, and speak or produce sounds like live char-acters. In addition, fires are lit automati-cally and new puppets appear on the stage until the end of the story.”

Heron’s book Belopoeica (Engines of War) purports to be based on a work by Ktesibius. His three books on mechan-ics survived in an Arabic translation. Mechanics is closely based on the work of Archimedes, presenting a wide range of engineering principles, including a theory of motion, a theory of balance, methods of lifting and transporting heavy objects with mechanical devices, and how to calculate the center of grav-ity for various simple shapes. Both books include Heron’s solution to the problem of two mean proportionals,

(a) (b)

Figure 11 Two representations of the steam boiler [14], [21] and physical reconstruction of the device by D. Kalligeropoulos and G. Pa-pamichail [21]: (a) a schematic from Heron’s writings and (b) a physical reconstruction of the device. The steam boiler, also known as the sphere of Aeolus, takes advantage of the steam pressure, transforming it into mechani-cal rotational force. The steam boiler is, of course, the ancestor of the steam engine. in the reproduction shown, the water is heated electrically and steam is passed to the sphere through pipes. The steam is released through a nozzle under pressure, creating a torque. The images are courtesy of the Thessaloniki Science Center and Technology Museum.

(a) (b)

Figure 10 (a) The force pump of Ktesibius and (b) a modern reproduction by D. Kria-ris [21]. A modern reproduction of the water pump delivered 1 m3 of water per hour with an efficiency of 80%. The reproduction was based on information found in the books by Philon, Heron, and Vitruvius [4]. The images are courtesy of the Thessaloniki Science Center and Technology Museum.

(a) (b)

Figure 9 (a) The water clock of Ktesibius and (b) a reproduction by D. Kriaris [21]. The imag-es are courtesy of the Thessaloniki Science Center and Technology Museum.


two quantities, x and y that satisfy / / / ,a x x y y b= = with a and b known.

This equation may be used to solve the problem of constructing a cube with double the volume of a given cube.

A catalog of 60 mechanical devices is attributed to Heron, not mythical but historical and indeed functional, as proven by their modern reproduc-tions. The function of his devices was based on scientific observations of nature. But unlike pure science, which involves theoretical understanding of the environment but often lacks imme-diate implementation, his observations were applied science brought to use. Heron himself, however, explained that some of these devices were not meant to cover any practical use but were created only for aesthetic plea-sure. Heron’s devices were explored and reproduced by Romans, Arabs, and Byzantines and served as the link uniting the technologies of the Ancient Greek and European Renaissance. Fig-ures 12 and 13 illustrate the oil press and a hoisting machine, which are two of Heron’s creations. Figure 14(a) shows the agiasterion ( ),oackavxftk ol which supplied an amount of holy water by throwing a coin. Figure 14(b) shows the water-powered mechanism for the automatic opening of the gate of

a temple. Figure 14(c) shows the diop-tra ( ),dkqrxtal a mechanism useful in geodesics, and Figure 14(d) shows the armonio ( ),oatnqokl a musical instru-ment operated by air pressure. Figure 11 depicts Heron’s steam boiler, and Figure 15 depicts the odometer, along with modern reproductions.

Mechanical achievements during those centuries were very significant. The world had one of the greatest geniuses, Archimedes, who devised remarkable weapons to protect his native Syracuse from Roman invasion and invented such basic mechanical devices as the screw, the pulley, and the lever. Alexandrian engineers, such as Ktesibius and Heron, invented a wealth of ingenious mechanical con-trivances including pumps, wind and hydraulic organs, compressed air engines, screw-cutting machines, and a steam turbine. Little practical use was found for these inventions, but the Alexandrian School marks an impor-tant transition from very simple mech-anisms to the more complex devices that properly deserve to be considered machines. In a sense, the School of Alexandria provided a starting point for modern mechanical practice.

GRECO-ROMAN PERIOd ANd ThE TRANSITION TO BYZANTIUMThe Roman Empire became the domi-nant force in the greater Mediterranean

region starting around 100 BC. The Romans were responsible, through the application and development of avail-able machines, for an important tech-nological transformation: the wide-spread introduction of rotary motion. Rotary motion was exemplified in the use of the treadmill for powering cranes and other heavy lifting opera-tions, the introduction of rotary water-raising devices for irrigation works (a scoop wheel powered by a treadmill), and the development of the water-wheel as a prime mover. Vitruvius, a Roman engineer, gave an account of watermills, and many were in opera-tion by the end of the Roman era.

The Roman Empire split into two parts around 350 AD, with Rome being the capital of the western part and the city of Byzantium, later called Constan-tinople (today’s Istanbul), the capital of the eastern part. The Eastern Roman Empire, also called the Byzantine

(a) (b)

Figure 12 (a) A representation of the oil press operated by an endless screw and (b) a modern reproduction by D. Kriaris [21]. The oil press, one of the oldest mechanisms and dated around 580 BC, capitalized on the ac-tion of an endless screw arrangement and the surface area of two plates to squeeze oil from olives. The screw rests on a hole at the center of the bottom fixed plate that allows the screw to move clockwise or counterclockwise to generate large pressure forces between the plates. The images are courtesy of the Thessaloniki Science Center and Technology Museum.

Figure 13 Hoisting machines: monokolos and dikolos, taken from [21]. The machines are also referenced in [22]. The photographs are courtesy of the Thessaloniki Science Center and Technology Museum.

(a) (b)

(c) (d)

Figure 14 Four inventions by Heron [10], [11], [19], [20], [21]. Heron continued the work of Ktesibius and Philon at the School of Alexan-dria and is righteously considered the spiritual ancestor of Leonardo Da Vinci for his famed constructions: (a) the agiasterion, which sup-plied an amount of holy water by throwing a coin, (b) the water mechanism for the auto-matic opening of the gate of a temple, (c) the dioptra, a mechanism useful in geodesics, and (d) the armonio, a musical instrument op-erated by air pressure. The images are cour-tesy of i.C. Lazos.


Empire, became distinctly Greek, as opposed to its western, Roman coun-terpart. For a variety of reasons, the level of knowledge declined much more rapidly in the Roman west than in the Greek east. Inventions in Byz-antium derive from Heron’s era, as shown in Figure 16, which depicts mechanical animals and birds of the Byzantine throne.

In the West, what survived dur-ing that era was mostly contained in handbooks as collections of known facts. Notable examples are Pliny’s Natural History (75 AD), Aulus Gellius’ Attic Nights (second century AD), and Slinus’ Collection of Remarkable Facts (third or fourth century AD).

In the East, by contrast, much more of the Greek science was preserved. The tradition of scholarship kept alive knowledge from the ancient scientific texts through scholarly discussions and the publication of commentaries in addition to original texts, even when the scholarly commentators themselves did not attempt to engage in original scientific research on their own account. Much of what is known today about Greek science is due to this tradition.

By the year 600 AD, preliminary efforts had already been made to turn part of the Greek corpus of scientific learning into Latin in the West and into Syrian in the East. Greek science

was converted first into the Arabic language in the ninth and tenth centu-ries and then once again from Arabic and Greek into Latin in the 12th and 13th centuries. Although the Greek corpus received major additions and modifications in the hands of Arab and Latin authors, it was still essen-tially the Greek learning that made its way to the Latin west.

After the fall of Alexandria to the Arabs (642 AD), knowledge of medi-cine, biology, astronomy, and mathe-matics spread through the Arab world and from about the middle of the ninth century, the Arabs produced excep-tional scholars, such as the astronomer and geometer Thabit ben Qurra and the polymath Al Kindi [16], [24].

dISCUSSION ANd CONCLUdING REMARKSThe central objective of this column was to describe technology and auton-omous mechanisms in the Mediter-ranean region and the ancient Greek world. The Mediterranean region was a very important hub, where the foun-dations of modern Western science and technology were laid. The foundations were very strong, based not only on ingenious solutions to specific prob-lems but also drawing on first prin-ciples that govern natural phenomena. Powerful mathematical concepts were also developed in the same geographi-

cal area during the same time frame. There is documentation and evidence of major inventions in areas such as architecture, metallurgy, marine engi-neering and ship building, hydraulics and hydraulic projects, construction engineering, musical instruments, and mechanisms using series of bolts and spur gears, to name a few [10], [21], that are not discussed in this article.

Ancient Greek technology was rather complex and contributed greatly to the development of Western civiliza-tion. It is wrong to claim that the ancient Greeks contributed only to philosophy, arts, and sciences exclusive of technol-ogy. This erroneous notion is partly due to Plato, who, in his fight against Archytas the Tarantine, expressed his passion for science, as opposed to tech-nology, by which he meant learning how to do or fix something [21].

It is true that, while most of the inventions discussed in this article originated in areas away from Athens, it is the Athenian philosophers who eventually recognized and supported the need for technology in parallel to, and in addition to, philosophy, arts, and sciences. More details may be found in [25].

An early version of this article was presented at the 2007 European Con-trol Conference in Kos, Greece and appeared in [26]; an extended version may be found in [27].

ACKNOwLEdGMENTSThe sources of the schematics and pictures used in this article, [10], [11], [17]–[21], [28], are acknowledged.

AUThOR INFORMATIONKimon P. Valavanis is the John Evans Professor and chair, Department of Elec-trical and Computer Engineering, Uni-versity of Denver, and director of the Un-manned Systems Research Institute. He is a graduate of the National Technical University of Athens and holds the M.S. (1984) and the Ph.D. (1986) from Rensse-laer Polytechnic Institute. His research interests are in the areas of unmanned systems, distributed intelligence sys-tems, robotics, and automation. He has

(a) (b)

Figure 15 (a) A representation of the odom-eter and (b) reconstructions of the odometer by D. Kriaris [21]. The odometer consists of an assembly of gears coupled through end-less screws that transform the wheel motion of a vehicle into units of length traveled. The reconstruction (b) was based on Heron’s writ-ings; it is fully autonomous and may be adapt-ed to fit any moving vehicle. The images are courtesy of the Thessaloniki Science Center and Technology Museum.

Figure 16 The mechanical animals and birds of the Byzantine throne are based on Heron’s inventions [10], [11], [19], [20]. it is speculated that these mechanisms moved hydraulically with water or steam pressure. The image is courtesy of i.C. Lazos.


authored, coauthored, and coedited 16 books and has published over 350 book chapters, technical journal/transaction, refereed conference papers, invited papers, and technical reports. He was editor-in-chief of IEEE Robotics and Au-tomation Magazine (1996–2005), and since 2006 he has been the editor-in-chief of the Journal of Intelligent and Robotic Sys-tems. He also served as a member of the IEEE Robotics and Automation Society Awards Committee for three years, and as cochair/chair of the Aerial Robotics and Unmanned Aerial Vehicles Tech-nical Committee. He served as Distin-guished Lecturer in the IEEE Robotics and Automation Society and he is a Senior Member of IEEE, a fellow of the American Association for the Advance-ment of Science, and a Fulbright Scholar.

George Vachtsevanos received a B.E.E. from the City College of New York in 1962, an M.E.E. from New York University in 1963, and a Ph.D. in electrical engineering from the City University of New York in 1970. He is currently professor emeritus at the Georgia Institute of Technology, after being professor of electrical and com-puter engineering from 1984 to 2007. He also directs the Intelligent Control Systems Laboratory, where faculty and students are conducting research in intelligent control, neurotechnol-ogy and cardiotechnology, fault di-agnosis and prognosis of large-scale dynamical systems, and control tech-nologies for unmanned aerial vehi-cles. He has published over 300 tech-nical papers and is the lead author of the book Intelligent Fault Diagnosis and Prognosis for Engineering Systems (Wi-ley, 2007). He was awarded the IEEE Control Systems Magazine Outstand-ing Paper Award for 2002–2003 (with L. Wills and B. Heck), the 2002–2003 Georgia Tech School of Electrical and Computer Engineering Distinguished Professor Award, and the 2003–2004 Georgia Institute of Technology Out-standing Interdisciplinary Activities Award. He is an IEEE Senior Member and a member of Eta Kappa Nu, Tau Beta Pi, and Sigma Xi.

Panos Antsaklis is the Brosey Pro-fessor of Electrical Engineering at the University of Notre Dame. He is a grad-uate of the National Technical Universi-ty of Athens, Greece, and holds an M.S. and Ph.D. from Brown University. His research addresses problems of control and automation and focuses on control systems that exhibit a high degree of autonomy. His recent research focuses on the design of cyberphysical sys-tems using passivity and dissipativity. He coauthored three research mono-graphs on discrete-event systems and on model-based control of networked systems, two graduate textbooks on lin-ear systems, and coedited six books on intelligent autonomous control, hybrid systems, and networked embedded control systems. He is a Fellow of IEEE, the International Federation of Auto-matic Control, and the American Asso-ciation for the Advancement of Science; the 2006 recipient of the Engineering Alumni Medal of Brown University; and a 2012 honorary doctorate recipient from the University of Lorraine, France. He is the editor-in-chief of IEEE Transac-tions on Automatic Control.

REFERENCES[1] O. Mayr, The Origins of Feedback Control. Cam-bridge, MA: MIT Press, 1970.[2] O. Mayr, “The origins of feedback control,” Sci. Amer., vol. 223, no. 4, pp. 110–118, 1970.[3] O. Mayr, Feedback Mechanisms in the Historical Collections of the National Museum of History and Technology. Washington, D.C.: Smithsonian Inst. Press, 1971.[4] S. Bennett, A History of Control Engineering, 1800–1930. London, U.K.: IEE Press and Peter Peregrinus Ltd, 1979.[5] S. Bennett, A History of Control Engineering 1930–1955. London, U.K.: IEE Press and Peter Peregrinus Ltd, 1993.[6] S. Bennett, “A brief history of automatic con-trol,” IEEE Control Syst. Mag., vol. 16, no. 3, pp. 17–25, 1996.[7] H. S. Black, “Inventing the negative feedback amplifier,” IEEE Spectr., vol. 14, no. 12, pp. 55–61, 1977.[8] H. W. Bode, “Feedback—The history of an idea,” Reprinted in Selected Papers on Mathemati-cal Trends in Control Theory, R. Bellman and R. Kalaba, Eds. New York: Dover Publications, 1960.[9] D. S. Bernstein and L. G. Bushnell, Eds., “His-tory of control,” Special Section, IEEE Control Syst. Mag., vol. 22, no. 2, pp. 21–68, Apr. 2002. [10] C. Lazos, Engineering and Technology in An-cient Greece, 11th ed. Athens, Greece: Aeolos Pub-lications, 1993.

[11] Proc. 1st Int. Conf. Ancient Greek Technology, Association for the Study of Ancient Greek Tech-nology and the Thessaloniki Technical Museum, Thessaloniki, Greece, 1998. [12] N. Karcanias, S. Vasileiadou, and D. Kallig-eropoulos, “The origin of systems concepts in ancient Greece,” in Proc. 2nd Int. Conf. Ancient Greek Technology, Athens, Greece, 17–21 Oct. 2005, pp. 732–738.[13] P. Antsaklis, G. Vachtsevanos, and M. Poly-carpou (Organizers), “Is feedback control a Med-iterranean invention?” Panel discussion on the Early History of Autonomous Mechanisms, in Proc. 13th Mediterranean Conf. Control Automation and IEEE Int. Symp. Intelligent Control, Limassol, Cyprus, June 27–29, 2005, TuPPE. [14] C. Marshall, Greek Science in Antiquity, New York: Collier Books-MacMillan, 1963.[15] G. E. R. Lloyd, Early Greek Science: Thales to Aristotle. New York: W. W. Norton, 1970.[16] G. E. R. Lloyd, Greek Science After Aristotle. New York: W. W. Norton, 1973.[17] P. Valavanis, Hysplex, the Starting Mechanism in Ancient Greek Stadia. A Contribution to Ancient Greek Technology. Berkeley, CA: California Univ. Press, 1999.[18] R. MacLeod, Ed., The Library of Alexandria: Centre of Learning in the Ancient World. London, U.K.: I. B. Tauris, 2004.[19] C. Lazos, The Adventure of Technology in An-cient Greece, 2nd ed. Athens, Greece: Aeolos Pub-lications, 1999.[20] C. Lazos, Archimedes, The Intelligent Engineer, 4th ed. Athens, Greece: Aeolos Publications, 1995.[21] Ancient Greek Technology. World Exhibition, EXPO 2000, by the Technological Museum of Thessaloniki and the General Secretariat of Re-search and Technology, Thessaloniki, Greece, 2000. [22] P. Vitruvius and M. H. Morgan, The Ten Books on Architecture. New York: Dover Publications, 1960.[23] “Heron of Alexandria, in Automatopoietica, D. Kalligeropoulos translation into modern Greek, commentary and original figures (in Greek), Foiniki, Athens, Greece, 1996. [24] A. Cameron, The Mediterranean World in Late An-tiquity AD 395–600, London, U.K.: Routledge, 1993.[25] D. Kalligeropoulos, Myth and History of An-cient Greek Technology and of Automata—Vol. A, Technology in Ancient Greek Myth (in Greek). Ath-ens, Greece: Kastaniotis Publications, 1999. [26] K. P. Valavanis, G. J. Vachtsevanos, and P. J. Antsaklis, “Technology and autonomous mechanisms in the Mediterranean: From ancient Greece to Byzantium,” in Proc. European Control Conf., Kos, Greece, July 2–5, 2007, pp. 263–270. [27] K. P. Valavanis, G. J. Vachtsevanos, and P. J. Antsaklis, “Technology and autonomous mechanisms in the Mediterranean: From ancient Greece to Byzantium,” ISIS Lab, Univ. Notre Dame, Tech. Rep. ISIS-2014-001, Notre Dame, In-diana, Jan. 2014.[28] C. Lazos, Hydraulic Instruments and Mecha-nisms in Egypt of the Ptolemies, 1st ed. Athens, Greece: Aeolos Publications, 1999.[29] D. Guedj, Le Théorème du Perroquet. Paris, France: Éditions du Seuil, 1998.

technology and autonomous mechanisms in the mediterranean: from ancient greece to byzantium...

Documents