What is automation? It usually conjures up science fiction imagery and limited ties to reality. But for a growing number of schools, labs, firms and professionals, automation is the word that gives professional passions meaning. This is because automation is, theoretically, a human helper. It is the codification (documentation and computing operationalization) of job tasks. Generally, automation describes a bundle of technologies that take routine, physical and thinking human tasks and turns them into algorithms and robotic procedures. The idea is that we can help with all of the work humans need to do, if we enlist intelligent computing technology. Importantly, as computing technologies progress, this notion of computing intelligence also becomes more and more like artificial intelligence.
Of course, the slight overreaction, here, is that our lives will be completely inundated with robots and we will not recognize our future in a world filled with artificially intelligent algorithms and robots. However, the reality is far more difficult to predict and likely not an immediate step into Star Trek, Star Wars, etc. It is more probable that a large portion of our job tasks will be “augmented” by the embedded technologies of automation, and that another portion of job tasks will grow from the automation economy. To an extent, some of these technologies will displace a number of job tasks and even whole jobs. Projections vary substantially on how this automation and job task replacement will pan out. However, it is possible that somewhere between 35 to 50 percent of job tasks will be augmented or automated by 2035. If this is even remotely true, and our projections are off by even 50 percent, then that is a transition of close to 200 million jobs globally in just over 15 years.
Why is all of this relevant to the built environment? Because real estate is notoriously slow to make changes, which is problematic for the sector’s long-run economic productivity. Buildings take a significant amount of time to create, and because they are the shell of all safe and healthy living, working and playing, there is a sense of responsibility to maintain rather than change. But research has shown that automation and its associated terms – namely, artificial intelligence, machine learning and deep learning – started to be developed as early as 1955. This means we have had some time to adjust to some of the concepts and ideas around automation. However, it seems that automation is still not widely accepted, and there are many barriers to its entry into the real estate industry.
In the MIT REI lab, we wondered why automation has been slow to take off in real estate, and subsequently devised a framework to highlight at least six reasons that outline its obstacles. These include limited product scalability, market-to-market idiosyncratic regulations, a limited number of lifecycle events that can synergize with automation events, numerous and complex, one-time engagement stakeholder experiences, liquidity risk in adopting automation technologies the market is not willing to consume, and limited technologies and business models to copy from other sectors. These obstacles are not impossible to overcome, but tackling them will require a different approach to dealing with industry-wide externalities.
Perhaps what is important, then, is understanding the phases of automation. There are three stages in our current outlook: recognition (data collection and tagging), sorting (machine and deep learning) and intelligence (unsupervised deep learning or so-called early general artificial intelligence). These stages are important, as they help us to gauge where our sector is in relation to others. Unlike financial services and life sciences, the real estate sector is still in the recognition stage with some firms engaging in the sorting stage.
The real estate sector has, in fact, made substantial strides in the past 20 years to cultivate “automation acumen” with the use of data and back-of-house automation processes. There are at least 150 data companies in the sector and there are a growing number of data integration companies. These companies are starting to pair thousands – even hundreds of thousands – of data sets to start to make inferences about past and future products and processes in the real estate sector. In addition, there is a growing number of robotic commercial offerings, with approximately 90 companies across the built environment. While some might find this surprising, Selina Short at EY does not. Instead, she sees firms developing “automation acumen” increasingly with their back of house operations. In these cases, starting with taxation and accounting tasks enables firms to understand the ethical and mindful codification process of automation.
With all of these factors in mind, consider whether automation is as far away as some skeptics would have us believe. Although we are still in the early stages, firms are taking note of what they are missing out on by not engaging in automation technologies and strategy. Fundamentally, it is economic growth stemming from efficiency, and potentially increased revenue from freeing up cognitive capacity to create new products. In any case, it is worth learning about what automation is, and how it impacts the real estate sector without the hype.