Commentaires

Enhancing the human experience: algorithms to drive individual workplace comfort

When it comes to workplace comfort, researchers at Purdue University are proving the theory that one size does not fit all.

05 octobre 2017

When it comes to workplace comfort, researchers at Purdue University are proving the theory that one size doesn’t fit all.

An ongoing study at the Purdue University Center for High Performance Buildings (CHPB), in partnership with JLL, is looking at the effect of customizable indoor environment conditions on employee productivity and satisfaction and building energy consumption. The goal of the two-year project, “Development of self-tuned indoor environments,” is to use measures of individual preferences to come up with smart building technology solutions.

“Essentially we are developing algorithms that can learn occupant preferences accurately and efficiently,” said Associate Professor of Civil Engineering Panagiota Karava, one of the researchers on the project.

Founded in 2013, the CHPB has quickly emerged as a leader in the smart building innovation space. Supported by partnerships with industry leaders, the CHPB takes a multi-disciplinary approach to its research projects, by bringing in experts in mechanical, civil and electrical engineering, as well as specialists in psychology and human behavior.

JLL has partnered with the CHPB on 13 research projects, including this one. The study is collecting data from over 200 participants in private and open-plan offices in an on-campus building that is a living laboratory. Each private office is equipped with dimmable electric lights, motorized shades and a Variable Air Volume system. In one sub-study, sensors measured temperature, light levels and occupant actions as participants were asked to engage in a regular 8-hour workday and interact with the customizable control systems from their desks. These systems were designed by the Purdue research team to be user-friendly and compatible with the building management system. Another group interacted with typical thermostat and lighting controls on the wall.

Results so far show that offices with easy-to-access controls resulted in a higher utilization of daylight and a corresponding decrease in energy consumption.

“People who could adjust their lighting and temperature of the room from their computers were more engaged than people who had to get up from their desks to adjust these levels,” Karava said. “By giving people the choice to use more daylight and feel more connected to the outdoor environment then you are optimizing productivity and using less energy.”

In addition to comfort, the study also monitored participant productivity levels through self-assessments and objective cognitive tests, including Choice RT and Spatial Stroop, that were conducted at different temperature and lighting conditions to measure the ability to translate a stimulus and the ability to inhibit interfering information. The study’s findings show that participants reported higher levels of self-evaluated productivity in customized environments and demonstrated better performance in cognitive tests (i.e. lower response time for accurate trials) when satisfied with the thermal and lighting conditions of their workspace.

People also acted fairly consistently over the duration of the four-month study, an encouraging sign that software could be designed to create specific profiles. The study’s aim, to create technological solutions to individualized preferences, relies heavily on learning algorithms, Karava said.

Faced with the challenge of collecting a large amount of data, the research team developed the novel idea of “clusters” – groups identified by similar general preferences and behavior in the workplace environment. With these cluster groups in place, unique profiles for new individuals could be measured using a fraction of the data points, and assigned values based on the percentage of preferences that aligned with each cluster.

“We have this clustering process that makes the algorithm more efficient, using less data and at the same time, creating this depth of prior knowledge so a new occupant does not have to be learned from scratch,” she said.

The preference profiles – or clusters – enable a faster path to identifying and meeting an occupant’s specific workplace needs. This process is enabling a shift to a more human-centric workplace environment, which has been slow in the real estate industry.

“The experience economy is fundamentally changing the real estate industry and its role in creating environments and measuring experiences that impact occupant satisfaction and business productivity,” said JLL’s President of Global Integrated Facilities Management Maureen Ehrenberg. “It’s not a one size fits all in customer experience, but one where we are creating, activating and measuring the impact of employee and customer experience.”

While the industry has been slow to adapt, JLL has emerged as a leader in this space with its recent global Human Experienceresearch report that delves deep into how dynamic experiences impact employee engagement, fulfillment and productivity in the workplace.

“We’re transforming a traditional services business into an experience business,” Ehrenberg said. “We’re leading the industry in delivering services focused on the convergence of facilities, workplace and technology. This approach enables us to deliver holistic, impactful experiences and to enhance productivity for our clients.”

Industry leaders like JLL have provided essential, real-world feedback to the project, Karava said.

“It’s great to see that we have common goals towards human-centered buildings,” she said.

JLL Energy & Sustainability Senior Director Leo O’Loughlin says that JLL’s partnership with Purdue fulfills a goal to provide clients with new insights that help them make informed, strategic business decisions.

“Our clients look to us to provide not only transparency on how their facilities are functioning now but to help match that with smart data on how to better achieve optimal, more efficient performance and enhance employee productivity. Research projects such as this help inform our clients on real-world scenarios to better guide them on their decisions.”

So far, the study has created a system with which to collect relevant data and an algorithm that can accurately create individualized profiles. The researchers’ plan for the study’s second year is to implement a prototype in actual office environments.

Vous aimez nos contenus ?