Designing and Managing Complex Systems

By David Moriarty

The systems that surround us are often multidimensional, and complex, consisting of a large collection of networked components with convoluted connections between them. Designing and managing such systems can be challenging, particularly in organizations. Designing and Managing Complex Systems introduces readers to the theory of complex systems, examining the role of human within larger systems, the factors that affect system performance, and how such systems can be optimized. The first section reviews the history of one particularly fruitful approach to complexity, one based on knowledge of the human nervous system. Next, the author discusses the current understanding of complex systems in a variety of domains including physical, biological, mechanical, and organizational. Within these chapters the author also introduces the idea that there are marked similarities in how complexity is successfully managed across these different domains and how the ideas from one domain can be useful in other domains. Next, these ideas are synthesized into a framework for successfully designing and managing complex systems. The fourth section focuses on case studies concerning failures and successes within complex systems.

• Provides an overview of the background and scope of complexity science
• Reviews current understanding of complex systems in a variety of domains (physical, biological, mechanical, and organizational)
• Introduces the idea of using successful techniques from one domain to help design and manage complex systems in other domains
• Includes case studies analysing failures and successes within complex systems

• Language: English
• Publisher: Academic Press
• Release date: Dec 1, 2022
• ISBN: 9780323916103

David Moriarty

David Moriarty initially trained as a doctor and during his medical training he undertook a BSc (Hons) in Neuroscience graduating with first class honors and having his research findings published in the British Journal of Neurosurgery. He graduated the Imperial College School of Medicine in 2004 after being awarded the Hawker Scholarship for highest preclinical examination results. His interest in neuroscience and the neurobiological basis of complex behavior led him to begin neurosurgical training before subsequently deciding to try an alternative career in aviation. He started flying for CityJet in 2007, became a captain in 2012 and subsequently became a training captain (a pilot that trains other pilots on the ground and in the cockpit). He is currently flying as a captain with easyJet. Much of the research into the complexity of both human performance and system design focuses on safety-critical industries such as nuclear, healthcare and aviation. After two years working in aviation, David became involved with Human Factors training, human factors being the scientific discipline that looks at how human performance and system design can be optimized. A year later, he became Chief Human Factors Instructor for CityJet. In order to improve his knowledge of human factors he completed a master’s degree in Human Factors and Safety Assessment in Aeronautics at Cranfield University with a specialist interest in Resilience Engineering and safety management in complex systems. Prior to starting the course he was awarded a Royal Aeronautical Society Centennial Scholarship and he graduated from Cranfield with the Course Director’s Prize for achieving the highest overall results in his class. He has published and presented work in the fields of neurosurgery, complex system management and aeronautical safety and also runs Zeroharm Solutions, a consultancy specializing in medical and aeronautical human factors. David is a member of the Royal Aeronautical Society, the Resilience Engineering Association and is part of SCiO, Systems and Cybernetics in Organizations. His first book, Practical Human Factors for Pilots was released on 15th January 2015 and is published by Academic Press, part of the Elsevier group. His approach to this current project is the same as his approach to his first book: to take the best science that is available across a wide range of disciplines and present it to the reader in an integrated, interesting and, most importantly, useful way.

Book preview

Part I

Cybernetics

Outline

Chapter 1.1. Control and communication

Chapter 1.1: Control and communication

Abstract

This chapter follows three scientists at different points in the early to mid-20th century as they attempt to understand how complex mechanical and biological systems are controlled and what internal and external channels of communication are significant in this control. The purpose is to give the reader an insight into one of the scientific disciplines that inform our understanding of complexity today, namely cybernetics. Aside from the history of this movement, the other aim of this chapter is to introduce the reader to the idea that rules of control and communication in one area of science can potentially be applied to understand control and communication in other areas.

Keywords

Communication; Consilience; Control; Cybernetics; Norbert wiener; Ross ashby; Stafford beer

I think the next century will be the century of complexity.

Stephen Hawking, physicist

The room was hexagonal, and its walls were paneled with darkly grained wood. Interspersed around the walls were displays of varying designs that could be controlled from any of the seven swivel chairs arranged in the center of the room. These chairs would not have looked out of place on the bridge of the USS Enterprise. Kirk's Enterprise, though, not Picard's. Low-backed and with the seat slightly tilted upward, they owed some of their design aesthetic to Charles Eames. However, the right armrest was widened and had been extended to house a control panel that would operate the screens. ¹ The room had been designed with a grand purpose in mind; it was to be the command-and-control center for the economic machinery of an entire country. It was the apotheosis of a bold field of scientific inquiry that had begun in the 1940s, a discipline that had been created and nurtured by some of the finest minds of the 20th century. It was also the grand hope of the new socialist regime in Chile and would potentially be the bedrock for a tremendous economic resurgence throughout the country. Then, on September 11, 1973, the socialist government of Chile was violently overthrown by General Augusto Pinochet. In La Moneda, the presidential palace, President Salvador Allende shot himself rather than be captured. Shortly after that, Pinochet's troops brok into the control room and destroyed it all.

In 1906, Norbert Wiener graduated from high school in Massachusetts and entered the prestigious Tufts University in downtown Boston. He was 11 years old. The newspaper, The World, hailed him as the youngest college man in the history of the United States and, with that, Wiener became the most famous child prodigy in America. His father was a Professor of Slavic Languages and had homeschooled his son according to a system of his own devising. Unlike many child prodigies who suddenly find that their youth is the only thing that separates their intellectual achievements from those of their older contemporaries at university, Wiener's academic progress was stellar. He went on to achieve his Ph.D. at 17, spending time studying under Bertrand Russell and G. H. Hardy at Cambridge, and the legendary mathematician David Hilbert at Gottingen. Although his primary focus was mathematics, Wiener also studied philosophy, biology, and engineering, subjects that would prove essential to much of his later work. In 1919, he became a lecturer at the Massachusetts Institute of Technology (MIT). He rose quickly to become a full Professor of Mathematics, a position he held for the next 40 years. ²

Short, a little rotund, and extremely myopic, Wiener was almost the perfect archetype of the absent-minded professor. This image was completed by thick-lensed tortoiseshell glasses and a Van Dyck beard. A notoriously bad driver, his MIT colleague Paul Samuelson remembered that his wife had expressly forbidden him to accept a lift home from work from Wiener despite them living in the same part of town. A late husband, she surmised, was better than a late husband. One particular story recounted by one of his colleagues concerned the day that the Wiener family was moving to a new house. Before he set off for work, Wiener's wife Margaret reminded him that they were moving today and that he should remember to go to their new address at the end of the day. Assuring her that he wouldn't forget, Wiener promptly forgot. Returning to his old address that evening, he was surprised to find that the house was empty and so asked a little girl who was playing in the street if she knew where the Wieners had moved to. Yes daddy, she replied. Mom said I was to stay here so I can bring you to our new house. ³

By the 1930s, Wiener had established himself as one of the academic stars in MIT's already glittering firmament. Although part of the mathematics department, he was a true polymath and could be reliably called on to contribute to cutting-edge research in many other scientific disciplines. He had a tradition of taking rambling Wiener-walks around the campus each day, stopping passing professors and students alike to expound on whatever idea or problem was concerning him at the time. On one occasion, he stopped a young physics student who was walking in the opposite direction. After their conversation, Wiener asked the young man what direction he had been walking when he had stopped him. You were going toward Building 8, the student replied. Thanks, Wiener said. That means I have already had my lunch. ²

The outbreak of World War II brought a fresh surge of interest in the practical applications of new scientific thinking. If World War I had been a war of attrition, the outcome of World War II would be decided according to which side could best harness cutting-edge science to gain a competitive advantage. The vice president of MIT, Vannevar Bush, had been selected by the US government to set up a new organization that would do just that—the National Defense Research Committee (NDRC). With full access to the country's top scientific minds, Bush was in a position to handpick leading scientists to work secretly on government defense projects. One such project was clearly going to require an immense effort and needed a first-rate mind to lead it. Fortunately, Bush knew the right man for the job, and Wiener was brought in to lead the project.

The challenge involved a new use for a new technology—radar. The military wanted to see if there was a way of using radar in conjunction with a statistical computation system to predict the future flight path of highly maneuverable warplanes. This information would then be instantly fed to the control unit of a bank of antiaircraft guns to shoot these aircraft from the sky automatically. The ranges over which these weapons were designed to work meant that you could not simply target an aircraft and shoot it down. By the time the shell reached the targeted area, the aircraft would have flown on a considerable distance. Instead, the guns had to fire shells to where the aircraft was going to be, taking into account the distance the shell would need to be fired and the expected flight path of the aircraft during that time. The challenge was considerable. It would have been difficult even if the pilots could be relied on it keep a constant course and speed. However, the pilots of that era were trained to undertake random evasive maneuvers if they were under fire. Wiener's task was to create a way of predicting how these evasive maneuvers would unfold during the course of an assault so that the antiaircraft weapons could be targeted not to the patch of sky where the aircraft was, but to the patch of sky where the aircraft would be. Much like the efforts to crack the German wartime codes underway at Bletchley Park in the United Kingdom, Wiener and his collaborators needed some way into the problem of how the complex decisions that the pilot must make under fire resulted in their choice of evasive action.

The answer came in the revelation that the options available to the pilot were more limited than they had first thought. ⁴ Traveling at high speed, the pilot can only make certain turns at specific rates. If he tries to turn faster than this or tries to turn in a direction that changes the aerodynamic load on the aircraft too much, he risks overstressing the aircraft or losing consciousness himself from the G-forces. By setting up a reciprocating information flow between the artillery system and the radar, the pilot's evasive actions following the first round of antiaircraft fire can be used to refine the second. The radar signals would update the statistical program to narrow down the possible areas where the target would be next. This process would continue until the target was shot down or it escaped.

Wiener had laid out the principles by which such a system could theoretically work. He now needed to embed those principles in a machine that would do the job. Firstly, the radar signals were fed into a makeshift machine that performed the calculations. Next, the results of these calculations would be fed directly to the gun turret that would then fire the shell. If the shell missed, the process would continue through another cycle, and the calculation would be able to factor in that the change in the flight path that had allowed the pilot to avoid the first shell left him with fewer potential flightpath changes available to him, which improved the chances that the second shell would hit the plane. Over subsequent cycles, the system was designed to progressively reduce the gap between where the shell exploded and where the aircraft was.

Sometimes, during the course of testing, Wiener and his colleagues noticed that the gun turret would exhibit some unusual behavior. Rather than progressively settling on the predicted target area, it would swing wildly to the limits of its allowed range of movement. A neuroscientist colleague of Wiener's, Arturo Rosenbluth, noted that this was similar to the tremor that a patient with Parkinson's disease experiences when they reach for an object. In these patients, it was common to see a slight but constant tremor in the hands when the patient was at rest. However, when the patients reached for an object or carried out a specific manual task, this tremor would become a lot more pronounced, almost to the extent that the arm and the wrist would move up and down to nearly their fullest extent. The neurological problem had been traced back to a fault in a set of circuits toward the back of the brain that would typically coordinate fine movement, i.e., the process of progressively reducing the difference in position between the desired object and the hand that is being extended to clasp it. Having realized that a mechanical circuit that carried out the same function would eliminate the gun turret's tremor, Wiener was able to refine the system, and trials showed that the mechanism could predict the path of a highly evasive aircraft up to a second before it got there. However, it soon became apparent that there was a problem. The calculations themselves were not an issue. The issue was the time available to complete them. Computers were a nascent technology in the 1940s, and even bringing the most advanced systems to bear on the problem couldn't help solve the equations quickly enough. Wiener had proved in principle that this sort of system was possible, but without the computing technology needed to put it into practice, the project was ultimately