Self-management (computer science)
Self-management in computer science refers to the development of autonomic systems that can operate and coordinate tasks autonomously, minimizing human intervention. These systems are modeled after the human autonomic nervous system, which automatically regulates vital functions. As technology has rapidly evolved, the complexity of computer systems has increased, creating challenges that self-managing systems aim to address.
Key objectives of self-management include self-configuration, self-healing, self-optimization, and self-protection, enabling systems to adapt, solve problems, manage resources, and safeguard against threats without human input. Prominent advocates of this technology, such as IBM, have significantly influenced its growth since the early 2000s. Examples of self-management include peer-to-peer file-sharing networks that utilize structured overlay networks for improved performance, and advanced technologies like self-driving cars and AI systems that dynamically adjust to their environments. Overall, self-management represents a significant step toward creating more efficient and resilient computing systems.
On this Page
Self-management (computer science)
In a self-managing computer system, all operations are run and coordinated autonomously without intervention in the process from humans. Such systems are also known as "autonomic systems," a phrase adapted from the medical term "autonomic nervous system," the area of the human body that automatically controls the respiratory system and the body’s heart rate. In theory, autonomic computer systems should function similarly.
As technology has grown rapidly, so has the complexity of computer hardware and software. In particular, the increase in bandwidth on the Internet has led to the development of complex applications such as online games and file-sharing programs, and mobile devices that serve the same functions as desktop computers. Different computing technologies and the need for each component to function properly as a whole system has created problems that experts hope will be solved by the emergence of self-managing computer systems.
Background
According to Forbes in 2024, the tablet computer market has grown significantly, in part because of Apple's release of its M4 iPad Pro, which was powered by Apple Intelligence. The number of smartphones also continued to grow, while other computing devices such as laptop and desktop computers, though not exhibiting the same type of growth, continue to be important consumer devices. Therefore, there is an increasing need, and shortfall, for computer technicians to manage these devices. However, autonomic computer systems could be a solution to bridge the gap between the growing number of devices and those able to manage them.
The biggest advocate and financial backer of autonomic computing in the past was IBM, which spearheaded the autonomic computing initiative (ACI) in 2001. IBM derived four goals that developers of autonomic computers should have: self-configuration, or the ability to optimize the various aspects of a system to help them function harmoniously; self-healing, or the ability to diagnose and solve problems automatically when errors occur in the system; self-optimization, or the ability to distribute resources to each component of a system so that each receives at least a minimum amount of resources; and self-protection, or the ability to ensure that outside effects such as viruses do not harm the system.
A typical example of the development of self-management in computers is peer-to-peer, or file-sharing, networks. Typically, applications that facilitate file sharing require a great deal of bandwidth to exchange information between one connection and another. Thus, there exists the possibility of bandwidth overload and subsequent failure to exchange the files. This failure was a common occurrence in early peer-to-peer applications, since they used what is called "random neighbor communication." By contrast, updated versions of these programs use what is called "structured overlay networks," which have the ability to use self-correcting measures if a failure is imminent.
Overview
The most fundamental idea related to autonomic computing is that machines can "learn," meaning they can gain knowledge and skills outside their programming and do not require a human to inject new skills into them. At the end of the twentieth century, IBM was the foremost technology company in its efforts to develop machine learning and associated technology. In the 1990s, IBM introduced Deep Blue, which was touted as a chess-playing machine that could defeat world champions. The company claimed that the machine could analyze over 200 million positions per second. In 1997, Deep Blue defeated Garry Kasparov, world chess champion, in a six-game contest. Considered by many to be the greatest chess player of all time, Kasparov was embarrassed by his loss to Deep Blue to such a degree that he accused IBM of intervening during critical moments of the matches, implying the company was cheating.
In 2011, IBM introduced "Watson," a computer system that used machine learning to defeat human contestants on the popular television quiz show Jeopardy!. Watson had the processing power of 500 gigabytes, meaning it could read and digest a million books in a matter of seconds. To enable this processing power, IBM engineers outfitted Watson with $3 million worth of hardware and uploaded nearly the entire contents of Wikipedia onto its servers. Watson was then matched with two of the most successful contestants ever on Jeopardy!. While Watson was able to defeat its human adversaries handily, its inability to interpret some questions as a human would was exposed. For example, when prompted with a question in the category of "US cities," Watson inexplicably answered "Toronto."
The most ambitious modern technological effort at autonomic computing was the self-driving car. While autonomous cars were conceived as early as the 1920s, they were not created until the 2010s. By 2022, seventeen US states allowed for the deployment of autonomous vehicles (AV) without a restriction on vehicle type. However, all but seven states required a human backup driver in AVs.
Other examples of autonomic computing included Amazon Web Services (AWS) Auto Scaling, which monitors applications and adjusts capacity to predict performance at a low cost. Scaling an application means giving it the ability to handle more work. In other words, it gives hardware and software more power when additional resources are needed. Doing this manually is time-consuming and expensive. Another example of autonomic computing is Google's DeepMind AI, a machine-learning system that adjusts cooling and ventilation based on what it learns about its environment.
Bibliography
"Autonomic Computing." DevX, 22 Aug. 2023, https://www.devx.com/terms/autonomic-computing/. Accessed 15 Nov. 2024.
"An Architectural Blueprint for Autonomic Computing." 3rd ed. IBM, June 2005. PDF file.
Bajarin, Tim. "Tablet Market Revival: Here's Why It's Growing Again." Forbes, 4 Sept. 2024, www.forbes.com/sites/timbajarin/2024/09/04/tablet-market-revival-heres-why-its-growing-again/. Accessed 15 Nov. 2024.
Cong-Vinh, Phan. Formal and Practical Aspects of Autonomic Computing and Networking: Specification, Development, and Verification. Hershey: Information Science Reference, 2012. Print.
Dargie, Waltenegu. Context-Aware Computing and Self-Managing Systems. Boca Raton: CRC, 2009. Print.
Kephart, Jeffrey O., and David M. Chess. "The Vision of Autonomic Computing." IEEE Computer Society, Jan. 2003. PDF file.
Lalanda, Philippe, Julie McCann, and Ada Diaconescu. Autonomic Computing: Principles, Design and Implementation. London: Springer, 2014. Print.
Michalski, Ryszard S., Jaime G. Carbonell, and Tom M. Mitchell. Machine Learning: An Artificial Intelligence Approach. New York: Springer, 2013. Print.
Nilsson, Nils J. Principles of Artificial Intelligence. San Francisco: Morgan Kaufmann, 2014. Print.
Zimmer, Ben. "Is It Time to Welcome Our New Computer Overlords?" Atlantic. Atlantic Monthly Group, 17 Feb. 2011. Web. 23 Aug. 2016.