Can artificial intelligence provide better processes and outcomes?

Oct 11
Machine Learning

Can artificial intelligence provide better processes and outcomes?

High tech, telecom, and financial services are the leading early adopters of Machine Learning and Artificial Intelligence. These industries are known for their willingness to invest and integrate new technologies to gain competitive advantage and internal process efficiencies. Now it’s the turn of the construction industry to show the world what we’re made of!

As an industry, we amass catastrophically large amounts of data and we can finally say we are starting to understand what to do with it.

Machine Learning allows us to generate progressive and independent thinking within the very machines and computers that drive our industry, in other words machines are learning to think for themselves. And it’s this vast amount of data we collect and analyse on a daily basis that is allowing the machines to learn and predict behaviours and patterns within occupied buildings as well as allowing designers to make informed choices within the design and prototype phases of a project.

The Machine Learning experience

Machine Learning is a branch of the more generic term Artificial Intelligence, and has many potential applications in construction for the design and operational stages of a project.  The area that shows the most potential to apply this technology today is in the operation of a building, in essence giving a building a brain.

Machines are very good at consuming and analysing large amounts of seamlessly unrelated data, and finding patterns in the chaos.  Once enough clean data has been collected it can be used for training the Machine Learning algorithm, allowing building systems to react to these patterns and make informed decisions.

In order for Machine Learning to operate effectively it needs to store, process and analyse huge volumes of data which once trained (a term used to teach the algorithm) it can drive a core set of algorithms that use neural networks with many hidden layers (also referred to as Deep Neural Networks (DNNs) to enable the learning, classification and prediction.

To put it in a more every day context, if we consider, when a child is born, it doesn’t know how to walk, talk, eat and so on, these are traits which are learned, usually through repetitive patterns which the human brain processes and interprets as common functions. Machine Learning operates in a very similar way to the neurological processes of the human brain, by looking at patterns within data it can make an informed decision on what will most likely happen next and will adjust accordingly. When a child eats sugar, it fires off an electrical signal to the brain that says ‘I like this’.  Over time when a child eats more sugar this emotion is fired again and that neurological connection becomes more burnt into the brain, the brain has essentially been ‘trained’.  This is very similar at a high level, how Machine Learning and neural networks work.  The main difference is that a machine does this on large existing data sets much quicker than a human can.

Another similar example, is if a computer programme is given thousands of images of frogs, and it is told thousands of times “this is a frog”, it will learn to recognise other frog pictures or objects that look like a frog.  Therefore, companies such as Google, who already have vast amount of data are leading the way in this field of Machine Learning.

In the same sense, if the programme sees thousands of architects selecting a particular type of door handle, for use on a particular door type in a particular building type, in a particular country, it can use this knowledge to suggest recommendations to future architects on which door handle to select automatically in their software, plus it will know why based on the reasoning of the previous architects. This is one of the reasons why many software companies are transitioning their platforms to the cloud so that they can start to connect the data from many different sources and connect the dots.

The SMART use of data

Similarly, the development of SMART buildings is increasing due to the availability of cheap sensors and IOT devices.  These are becoming common place in our homes, with many everyday objects now available as SMART products. Many buildings already have sensors that measure; light, heat, energy consumption, footfall, movement, capacity and much more, however many of these systems are proprietary closed systems that don’t talk to each other.  This is changing with many new systems been able share data in an Open Standard format or via an API interface.

Once this data is connected and able to be stored we can layer one of the many Machine Learning algorithm’s over the top to allow us to predict the behaviours of the building and its occupants, not only for efficiencies in operating and maintenance, but to inform the way we design future buildings to optimise space, flow, materials selection etc.

Many clients are now asking to know when an asset is likely to breakdown so it can be replaced, rather than wait until it breaks or fails.  Once we have enough data, we can use Linear Regression algorithm on the data to predict asset failure.

Machine learning is also used for more every day tasks, such as providing more relevant search results, by using previous user search trends to influence the results.  This has been used for years by companies such as Netflix or Facebook who use Machine Learning to push news stories to your Facebook feed.  Through the BIM object provider, bimstore they are already working on implementing such technology into their current platforms, and are now able to provide manufacturers with insights into not only what specifiers have selected, but what may have influenced their decision.

We know that construction is still years behind in the Artificial Intelligence field, the computers in a modern car, or aeroplane collect and analysing vast amounts of data every second.  Companies like Tesla are using this technology analysis our driving patterns, to diagnose any problems even before they happen, or make recommendations on when to change the tyres and which tyres to select based on predicted driving patterns, distance travelled and road surfaces and that’s only the very tip of the iceberg for Machine Learning capabilities.

With many Buildings now been designed, delivered and operated, using digital processes and technologies such as Building Information Modelling (BIM), Machine Learning is also starting to appear in many of the tools that architects and engineers use daily.  Companies such as Autodesk are investing heavily and their technology previews such as Project Dreamcatcher and Project Fractal, are hinting at the ultimate future where machines will be able to ‘design’.

For the past few years, there has been a strong message that in the future we will not be telling the computer “what to do” as we do now, and instead will be telling the computer “what we want to achieve” and letting the computer find the optimal solution, and the optimal design.

Autodesk are also looking at how Machine Learning can make construction sites a safer place and have recently acquired Smartvid.io, who have developed intelligent video and photo tagging software which allows cameras to highlight potential hazards and issues on a construction site before they happen.  For example, making sure that everyone on site is wearing the correct protective equipment, or that only approved site operatives can operate machinery.

The rapid interest and development around machine learning in the last couple of years has been caused by a perfect storm of emerging technology’s maturing across all industries.   Data is becoming more accessible through cheap internet of things devices, open standards and public API’s allow devices and services to talk to each other and combined with ease of access to cloud storage and infinite computing, where we can now collect, store, and analyse vast amounts of data, something which is important as the more data available the more accurate the Machine Learning algorithms are becoming.

Data is the new oil!

The information we can derive from data and the patterns its presents on our actions, selections, behaviour and choices means whomever can own, control and interpret this data has immense power – data has become the modern-day oil with large organisations such as Facebook, Google etc. all competing for your personal data.

And with this comes associated high risk, which is why the new data protection GDPR regulations are changing to adjust to this new threat, and in Digital Construction we have standards to protect the data and its use in specific standards such as PAS1192-5:2015.

This new power gained from data and the intelligent use and interpretations of it is changing the way our industry operates, for many it has already become the norm.

Adam Ward, Technology Director at BIM Technologies