The future: Fourth revolution of mankind
Authors
![Andrew in an open collar blue shirt with arms folded, smiling to camera in front of the office window](https://cdn.cundall.com/uploads/images/People/London/Andrew-Jackson/_contact_thumbnail/415148/Andrew-Jackson_02.webp?v=1707220594%2C0.6332%2C0.3499)
Andrew Jackson
View bioWe are currently living through the fourth great revolution of humankind.
In the wake of the cognitive revolution, the agricultural revolution, and the industrial revolution, the digital revolution is now driving fundamental changes in almost every aspect of human existence at a rate never seen before. But what does this mean for the future? Looking at the recent past may provide some clues.
In the aftermath of the dot com bubble in 2004, only two of the top 10 global companies by market capitalisation were tech companies: Intel and Microsoft. At this time, these companies followed a fairly traditional business model of making and selling products - software in the case of Microsoft and hardware for Intel. In the aftermath of the 2007-2008 financial crisis, oil companies were resurgent and Chinese companies came to the fore as US companies struggled. Microsoft was the only tech company remaining in the top 10[1].
By 2019 the picture was radically different: seven of the top 10 companies were technology companies, with no oil companies listed in the top 10. These tech giants also trade in completely new business sectors structured around data and services. Even Microsoft, which traditionally earned revenue from selling software has transitioned to 'software as a service' (SAAS) model. The single biggest contributor to Microsoft’s revenue is now cloud computing services[2].
There are two big headlines to take away from this: the decline of the oil-based economy and the beginning of the tech era.
But what will be the impact of this upon the construction sector? Let’s consider how the advent of artificial Intelligence, big Data and the transition away from fossil fuels might change the built environment and the way building designers work.
Cognitive buildings
Most of us have heard the term ‘smart buildings' but defining this is still not comfortable for many.
A typical building today is usually controlled with a varying degree of automation by several different sub-systems which might include the Building Management System (BMS), lighting control system, access control, security, meeting room bookings etc. These systems are often able to communicate with each other to a limited extent but are largely standalone.
In smart buildings, the various sub-systems are interconnected by a systems integration layer which allows them to freely communicate and interact. Further systems integration might also allow a building to interact with both its users, via smartphones and user tracking, plus systems surrounding a building such as smart utility networks and local public transport networks.
This creates completely new opportunities to enhance user experience, improve building operational performance and reduce wastage. For example, users may be notified when it is time to leave to catch their train home; spaces which are not required can be put into setback mode; catering provisions can be matched exactly to the number of occupants on a given day.
However, despite this increasing level of integration, most of the automation in smart buildings still relies upon pre-programmed if-this-then-that logic (IFTTT). The problem with this is that it relies largely upon a building designer’s expectations of how a building might be operated and doesn’t address changes that might happen over the lifecycle of a building – automation code is rarely revisited.
The future introduction of artificial intelligence (AI) will create a completely new category of ‘cognitive’ buildings. These buildings will be operated by an algorithm which is given no fixed programming but is simply an optimisation parameter and some simple boundaries or limits which must always be obeyed. Optimisation parameters might simply be “use least energy” or “maximise occupant satisfaction”.
The algorithm will then experiment to achieve the best outcome and continually optimise the operation of the building, such that it evolves over the course of its life. Effectively, buildings will be continuously commissioned to ensure they operate in the most efficient way possible whilst also adapting to feedback, always ensuring the best possible user experience.
So why aren’t these possibilities already being deployed in the built environment?
One difficulty is that AI is most easily applied to Markov systems. In a Markov system, the future state is only dependent upon the current state and any subsequent inputs - they are therefore independent of preceding events. However, due to transient and time-dependant effects, buildings are highly non-Markovian in their behaviour which makes the application of reinforcement learning challenging.
Secondly, most AI applications rely on a ‘training set’ of data upon which the algorithm can do most of the required learning outside of the real-world environment. For example, the algorithm needs to learn that opening the windows in the middle of winter increases energy use while reducing occupant satisfaction and so is a bad move. Clearly the AI needs to 'learn' this before it is let loose on a real building! In practical terms, this training set could be a digital model of a proposed building, however, buildings are extremely complex systems, and a suitably developed model is often not available.
Ongoing advancements in AI, dynamic simulation modelling (DSM) and digital twin will help to overcome these challenges with more sophisticated algorithms and better training sets available earlier in the design process than ever before.
One of the major benefits of this technology is that it could also be applied to existing buildings with relatively minor physical interventions. This is potentially key to meeting the UK government’s net zero pledges given that 80% of the 2050 building stock has already been built[4].
The potential rewards of simply operating buildings in the most efficient way possible are huge.
This article is the first in a two-part series. The second article will be published shortly.
[1] Ref: https://www.visualcapitalist.com/a-visual-history-of-the-largest-companies-by-market-cap-1999-today/
[2] Ref: https://techbehemoths.com/blog/how-microsoft-makes-billions
[3} Ref: A Visual History of the Largest Companies by Market Cap (1999-Today) (visualcapitalist.com)
[4} Ref: UK Green Building Council, Climate Change