Pre- and Post-Occupancy Evaluations Go Digital

Despite the effort that goes into design, architects often remain—sometimes willingly—in the dark about how their buildings serve their occupants. Increasingly, data-driven clients want to verify the design goals of their completed projects.

And, as revealing as a post-occupancy evaluation can be, the pre-occupancy evaluation is equally as critical. Documenting the “before” state of an environment provides a solid benchmark from which to calculate the impact of the design. Otherwise, says Mardelle Shepley, FAIA, a professor and chair of the department of design and environmental analysis at Cornell University, “you cannot determine if a transition has been made.”

While experience, best practices, and intuition continue to guide design decisions, and surveys, observations, and building monitoring data inform occupancy evaluations, architects can now leverage technology to collect user feedback and performance metrics en masse, analyze the information, and distill the results, improving the entire process of design.

Tech for Every Project
With recent advancements made in computing, data-seeking firms now have ready access to apps and machine learning tools.

Global firm IA Interior Architects and Portland, Ore.–based ZGF Architects have each developed custom iPad apps for the pre-occupancy evaluation process. IA’s tool, called IA Survey, lists pre-determined variables—for example, the number of occupants in the room, whether they’re working solo or collaborating, and if they are sitting, standing, or working with a physical object at a table, such as a computer—for observers to quickly enter data, snap pictures, and take notes.

IA designed the app in-house and then hired a software developer to build it. Already in its third year of use, the investment has paid off: Observation teams have cut their data collection time in half. The firm also shares the app with their clients for their active participation in the evaluation process.

Surveying, long a mainstay of occupancy studies, is also enjoying an upgrade through technology. To automate the process of surveying employee workstations during pre-occupancy evaluations, ZGF is experimenting with machine learning, the field of computer science that aims to develop programs that can learn and adapt over time through observations, artificial intelligence, and real-world interactions with minimal human direction. Instead of having an observer tally desktop equipment and configurations, ZGF would use a machine learning routine to identify items from photographs of each desk.

For firms without extensive in-house development resources, startup companies are providing machine learning services in more accessible packages. Prism Skylabs, based in San Francisco, provides a cloud-based image recognition service that can be trained to count objects and faces (which the software blurs to protect privacy). Firms can upload their images to Prism Vision platform and let its machine learning routines run their course.

Not surprisingly, the juggernauts in tech also offer solutions geared toward business users. The Amazon SageMaker platform—by, yes, that Amazon—claims to make the development of machine learning applications as easy as creating a website. Likewise, the AWS (Amazon Web Services) DeepLens camera, due in June, utilizes deep learning technology to analyze and react to real-time video footage.

Microsoft’s Azure and Google’s Cloud Machine Learning (ML) Engine platforms are also poised to make machine learning much more accessible. Azure’s Machine Learning Studio is a browser-based authoring environment that provides access to sophisticated tools without requiring any programming knowledge from users. Though not as user friendly as Azure, Google offers a free crash course on the basics of machine learning using its platform.

Similarly, the increasing prevalence of the Internet of Things (IoT) and capabilities of sensor technology provide more opportunities for information gathering in pre- and post-occupancy evaluations. While issues of data ownership may preclude architecture firms from installing their own sensors into their clients’ spaces, those clients might be willing to share any data they collect. Though IA is not receiving much sensor data from clients yet, due largely to the limited number of installed sensors to date, director of design intelligence Guy Messick expects that to change soon: “It’s definitely coming and there will be a demand for it.”

Sarah Bird, a workplace strategist and change management consultant at IA’s Chicago office, stresses the need to balance input from humans and machines. “Technology provides a lot of ‘what’ but you don’t always get the ‘why,’ ” she says. “You need to ask the questions to get the answers to the ‘why’ questions.”

^{For a proposed corporate campus of up to 15,000 end users, ZGF ran multiple circulation simulations to assess placement of programmatic elements, such as cafeterias, as well as furniture and pathways.}

The Why and the How
Before unleashing machines into their clients’ spaces, firms must conduct due diligence on what constitutes occupancy evaluations. Cornell’s Shepley recommends that architects first read up on formal research methodologies, affiliate with academic institutions, and hire employees with research backgrounds to develop the practice.

International firm CannonDesign staffs evaluation teams with employees well-versed in talking to user groups and who “know how to ask the right questions,” says PJ Glasco, AIA, a principal and healthcare planner and designer at CannonDesign’s Houston office. The firm also partners newer staff with experienced designers to learn firsthand what does and doesn’t work.

Firms interested in testing new evaluation technologies can turn their own workplace into a sandbox. ZGF is currently experimenting with infrared cameras to measure pedestrian flow in its Seattle office. A series of cameras mounted along a stair tracks the stair’s usage rate and the speed of movement among users. This method allows for a continuous occupancy evaluation of the space, says Paul Diaz, a building performance specialist based in ZGF’s Seattle office. “A building is really a fluid dynamics problem,” adds Dane Stokes, a design technology specialist at ZGF. “When you consider it that way, snapshots are not very useful.”

^{Infrared cameras track user speed, direction, peak use times, and collision rates on stairs in ZGF’s Seattle office.}

Streamlining the Feed
Garnering client buy-in is as essential as understanding the protocols and objectives of occupancy studies. Surveying users and observing the utilization of space are time-intensive endeavors, as are analyzing and reporting on the information. Unless the project is of a massive scale, the additional fee isn’t likely to pay immediate dividends to the owner— or architect. However, technology is helping to reduce the demands on architects’ workloads and open occupancy studies to smaller projects.

Sensors are turning data collection into a remote and automated process, where occupancy and building conditions can be captured continually and fed directly into a database. Mobile observation apps such as KieranTimberlake’s Roast, where occupants can opt to answer quick check-ins about their space conditions and comfort level, can also provide invaluable localized data with minimal disruption to the employees’ workday.

Closing the Loop
With potentially massive amounts of data in hand, thanks to these digital collection tools, architects can again turn to computing to disinter and present the takeaways.

For example, IA’s Survey app exports its data to Microsoft Excel, which then links up to Microsoft Power BI (Business Intelligence) to conduct the analysis. Using a dashboard that automatically tallies daily office and conference room occupancy, teams can quickly organize and share the data internally and with clients via an online portal. Tableau Software’s namesake data visualization tool can also help parse numbers into digestible bites for users. At CannonDesign, after the team leading an evaluation shares the results of an occupancy study with the client, it will present the findings to the firm’s designers. With the client’s permission, the team will also share the findings in an in-house pinup presentation and in marketing case studies.

“Clients are looking for evidence for design decisions,” says Tim Deak, a workplace strategist at ZGF. With the digital tools available now, pre- and post-occupancy evaluations can offer these takeaways with fewer intrusions on everyone’s workday. This in turn can strengthen the case for investing in architecture.

Note: This article appears under the heading “Digitizing the Design Feedback Loop” in the May 2018 issue of ARCHITECT. This article has been updated since first publication to correct two attributions.