News

Using AI to make our workplaces work better

Trish Sunga Trish Sunga Australasia Press Office,Sydney
13 August 2020

New research into how work environments affect productivity gives clear clues to managers and designers organising ‘new normal’ office arrangements, or planning fit-outs or refurbs. 

So too, people setting up more permanent home offices. 

The research between Arup and RMIT University is attracting world-wide attention, especially as workplaces adapt to covid-19 restrictions. 

The combined team of artificial intelligence specialists, psychologists and engineers examined the effects of work environments on productivity indicators such as concentration, comfort and activity. 

Pioneering an AI sensor-based system, the team monitored noise levels, indoor temperature, air quality, humidity, air pressure and electromagnetic fields: they also surveyed people’s views on disruptive influences, including meetings, over the course of several months.

The project used advanced machine learning methodologies to start to quantify the relationship between physical environments and worker wellbeing. ” Dr Mohammad Saiedur Rahaman Dr Mohammad Saiedur Rahaman Research Fellow, RMIT

The focus was on activity-based and mobile workplaces – where people can choose however, and wherever, they feel most comfortable, including alternating between the office and home – researching two large open-plan workplaces with different designs and layouts.

Dr Rahaman explained: “The research team used the data and survey information to train machine learning algorithms that could identify patterns in perceived concentration and activity. 

“Analysis of the patterns then led to development of solutions for making these workspaces function best for people.”

The research provides a mechanism through which we can look at how effective a fit-out design is, so that moving forward, we can be smarter in how we design buildings. ”

Tim Rawling Engineer

Arup engineer Tim Rawling described the study as ‘a world leading piece of research that opens the door to a broad range of applications of this and similar algorithms to building design’.

“We want to make workspaces that are better for the people who use them,” said Tim. 

“The research can help designers and organisations better understand the emotions, efficiency and team dynamics of employees, which could lead towards more overall organisational success.”

This study opens up an exciting opportunity to be able to predict the future performance of a workplace’s design right from the beginning, before breaking ground. ”

Shaw Kudo Engineer

Some findings were surprising. Arup engineer Shaw Kudo explained: “While we anticipated noise and lack of privacy as downsides for some people, one of the key factors causing a negative work experience was the availability of seating. 

“Regardless of their demographic profile, many people had concentration difficulties if they couldn’t sit in their favourite spot, which might be characterised by proximity to team mates, natural light, facilities, temperature comfort and so on.” 

The RMIT-Arup joint project team: (left to right) Yongli Ren, Dr Mohammad Saiedur Rahaman, Shaw Kudo, Tim Rawling, Flora Salim and Jonathan Liono.

High CO2 levels were also a barrier in people’s ability to concentrate. 

The research also showed the influence of non-environmental factors: there was a clear correlation between people having a lot of meetings (both formal and informal) and high levels of stress, and the impact of tiredness. Overall, better perceived concentration levels were recorded in the morning for the trial group in Melbourne.  

As covid-19 concerns and restrictions continue, many people will alternate between home offices and their traditional places of work. 

“There are far more variables influencing concentration when working from home,” said Tim. “People should deliberately notice their environment and what they need to be productive, and then address those potential barriers to their productivity, concentration, health and wellbeing.

“Regarding tiredness, being home-based can potentially help, as it becomes easier (for some) and more accepted to work adjusted/flexible hours that better suit sleep cycles.” 

Shaw continued: “Employers should think about employees’ work from home set-ups and how they can help bring the best out of each person during these difficult times.

“Extended working from home experiences are likely to make people even more aware of what does, and doesn’t, work for them when they return to an office environment. 

“Physical distancing and hygiene rules mean different workplace designs, layouts and available facilities – all of which could exaggerate the productivity downsides of not accessing preferred seating.”

The study ‘An Ambient-Physical System to Infer Concentration in Open-plan Workplace’ was published in the IEEE Internet of Things Journal (DOI: 10.1109/JIOT.2020.2996219).