Last edited 17 May 2019

Artificial intelligence

Electrons.jpg

Humans and animals display varying levels of natural intelligence. Thanks to integrated software programmes, machines can be enabled to perform some tasks intelligently, thinking and working in a similar way to natural intelligence. This ability of machines to act in this way is called artificial intelligence (AI) or machine intelligence.

A branch of computer science, AI is a machine's ability to imitate intelligent behaviour and learn from its operational history or experience. Typical examples of AI include speech and facial recognition, learning, predicting and problem solving. Artificial intelligence will also provide the analytical power behind self-driving vehicles navigating to their destination with human passengers while avoiding accidents.

[edit] Learning with AI

As a key element of AI, machines can learn, draw conclusions and act independently of humans if they have sufficient information about what is going on in their particular area of activity. The spaceship computer HAL in Stanley Kubrick's '2001 A Space Odyssey' displayed a particularly malevolent form of AI, thinking and acting against the interests of its human programmers.

AI allows a machine to learn and identify patterns and trends from gigantic quantities of real-world data (big data) without human supervision. Computers are now better than humans at playing chess - and can teach themselves how to play - and can also, for example, spot the danger signs that humans might miss on cancer scans.

Google's DeepMind research project into AI aims to develop programmes; "...that can learn to solve any complex problem without needing to be taught how". An unbeatable chess-playing computer is a typical example. The firm claimed that if the project is successful, it will be of a huge benefit, tackling some of the world's most pressing challenges, from climate change to improved healthcare.

[edit] Examples of AI in the construction industry

Using AI, the effects of billion-pound infrastructure will be predictable before planning and design, making infrastructure delivery more effective and economical. AI-empowered autonomous equipment can navigate within its surroundings without human assistance. This type of equipment can survey a proposed construction site, and create 3D maps and plans from the information it has gathered.

AI can advise engineers on the best methods to use on projects. For a super-tall building, AI systems could indicate the most optimal and economic structure.

One of the most groundbreaking applications to have come into the building industry in the past few years is Building Information Modelling (BIM). This allows the creation of a virtual model holding information that may encompass all aspects of the construction project across the building's lifecycle. Architects, engineers and others who have access to the BIM model can amend, update and consider the effect of various changes to the project under a wide range of parameters, including cost, programme, performance, buildability or clash detection. When AI processes are merged into BIM, the result will be the creation of an all-encompassing tool that gives full and intelligent control and prediction of planning, design and construction.

[edit] Related articles on Designing for Buildings Wiki