Think smarter

How to overcome bias in lift tech. By Tom Harmsworth


How to overcome bias in lift tech. By Tom Harmsworth

In Professor Jennifer Eberhardt’s book, Biased, she describes how a Black police officer pulled his gun on a man in New York only to realise he was looking at his own reflection. It is a striking anecdote, and one that demonstrates how insidious and powerful unconscious bias can be. Eberhardt is one of the world’s leading experts on unconscious racial bias, and shows in her book how this affects every sector of society, leading to enormous disparities. Only last month [at time of writing], Alexandra Wilson, a barrister specialising in criminal and family cases, was mistaken for a defendant three times in a single day.

Bias (despite what some like to claim) is not always, or even often, deliberate. It is an embedded trait of human cognitive function and one that plays an important role in synthesising the huge amount of data we have to process every day. The truth is that often we just don’t appreciate how vulnerable we are to certain narratives, how under the influence we are to the things we’re told and the things we see. And we don’t realise how this is reflected in the way we build and relate to the world around us.

This aspect of human nature has bled into the world of technology. In fact, there are few people now who would contest the fact that bias exists in the world of tech. Predominantly white male developers have created technology that responds principally to people like them, and so unfairly discriminates against women and people of colour. A 2018 survey showed that around three quarters of all technical jobs were held by men, and the needle hasn’t moved much since.

This is reflected in social media algorithms, facial recognition technology, and voice assistants, all of them increasingly common features of modern life, and all of them coming into the orbit of lift tech. ‘Smart’ lifts, some of which can be found in residential compounds and other buildings in Shanghai, for example, feature facial recognition technology as a security measure, to alert police to intruders or to alert rescue teams if someone gets stuck. This has yet to really take off in the UK, but AFR has already been used to make arrests in Britain.

Voice-controlled lifts, too, are coming into fashion. Soon, it might be common to tell your lift which floor you’d like to be taken to, rather than to select it by hand. And these trends are being accelerated by the pandemic, which is prompting people to come up with ways of maintaining full functionality but hands-free. Naturally this means prioritising technology like AFR and voice assistants.

But this trend towards AFR and voice assistant tech is also being propelled by the changing populations of buildings. The days when one mega-company would own the entire building are mostly over. Now, most buildings are shared; co-working spaces, despite the pandemic, are also on the rise. What this means is that security doesn’t begin and end with the entrance to the building, and the use of AFR and voice assistants may be increased in lifts specifically as a safety measure.

We can do something about unconscious bias. Clearly, we need to include as diverse a range of people in the tech-development process as possible, and from the very start, in order to make sure that the final product is not unfairly prejudiced against one group or more. This is a lengthy project, since it goes well beyond the world of tech and into wider society and culture. But we can nonetheless, within our own organisations, strive to be inclusive when we create new technology.

And for this to last, we will need to go beyond inclusion to education. The ‘why’ is as important as the ‘how’. We need to understand the flaws in the way we think—the kind of flaws that Professor Eberhardt illuminates—so that we do not keep making the same mistakes we’ve made in the past, and so that we approach our work in full knowledge that we are prone to bias and that the flaws in our psychology can creep into the technology we fashion for ourselves, whether that’s in the lift tech space in elsewhere in our world.

In these unpredictable times, we can predict that the lifts of the near-future will rely even more heavily on technology than they do now. And that means that the clock is ticking, and we need to iron out those deficiencies and vulnerabilities in our tech. That needs to be worked into the culture of tech, so that the technology of the future, whatever it might be, does not have the flaws that the technology of the past (and present) had. The first step is to have no illusions. But then we need to work hard on technological inclusion and education—and show this space to be as forward-thinking as it can be.

Tom Harmsworth is managing director UK for WeMaintain, the tech-enabled lift company radically transforming the market. It gives building managers a service that combines the technical skill of engineers with the agility and predictive power of its proprietary technology.

WeMaintain takes care of the invisible—but essential—operations of a building. By putting quality at the forefront of regulated lift maintenance, it allows managers to optimise their services and provide the best experience to their occupants.

For more information, please see www.wemaintain.com/en-uk