This is the second post in a series about the LMN Office Renovation focusing on the technologies we leveraged to make a better space in which to work. The LMN Office Renovation offered a chance to take an existing International Style building from the 1950’s and, through the use of modern technology and control systems, create an environment that significantly increased occupant comfort and well-being. As mentioned in the previous post on Daylighting the starting point for the endeavor was to determine the areas of greatest dissatisfaction. Through interviews and occupant surveys we were able to isolate four main issues:
The first three we discussed previously. In this post we will explore the nebulous topic of human comfort.
The top three comfort goals, while important, were not the only issues that occupants had with the space. The HVAC system is as old as the building. The controls are not always automated or tied to thermostats. There are two primary systems that were designed for the initial layout of the building. There is a perimeter system that is controlled by mixing valves that are adjusted by hand. These were intended for individual control when the building was ringed by private offices. There is also a central system that is controlled by thermostats and which served the conference rooms, kitchens, restrooms, lobbies and other interior rooms. When LMN decided to move to an open floor plan to allow better light penetration, communication, access to views, etc., the initial logic from the building’s original HVAC design was challenged. In order to monitor and control temperature across the open floor plate, LMN had to come up with a way of gathering data in multiple parts of the office in an objective manner in real-time. But because occupancy comfort can be nuanced, LMN decided to gather data by monitoring more than just temperature and to log that data in order to track changes over time. To do this LMN, is using an Arduino® circuit board and a host of sensors, to make office occupancy comfort sensor pods. The Pods The sensor pods sense a wide range of data: temperature, barometric pressure, sound, light, humidity, carbon dioxide, and motion. Prototype Pod
LMN can make the sensor pods quickly and affordably, which allows us to spread several throughout the office at typical workspaces. Each pod is connected to a network and uploaded instantly to Xively®, an online service offering a database and interface for the results. It is a website that allows LMN to share this database with the public. LMN can also connect easily to Arduinos within the Xively library, and this library allows LMN to not only receive data, but also send data to the Arduino to trigger actuators: Xively dashboard
Since LMN fabricated the post-occupancy sensors using various hardware from different manufacturers and our own code, there is a need to confirm the accuracy of our readings. For this, LMN ordered a HOBO® sensor for benchmarking the custom setup. HOBO® is a leader in occupancy comfort data logging, and their hardware is top notch (as well as top dollar). With one HOBO® sensor, LMN can benchmark a fleet of the Arduino sensors and verify their accuracy. LMN also ordered a Sound Level Meter so that we may calibrate the Arduino sound sensor. Some may ask why we would build our own sensors when there are good products already on the market. The reasons are many but primarily it is to learn more about the issues we are studying by getting our hands dirty. At LMN we believe that making is an essential part of what we do. It’s analogous to sketching versus photographing. When you sketch something you have to pay much more attention to the details and to the subject’s composition. So it is with the making of the sensor pods. The understanding that is generated by creating the sensor pods is invaluable to LMN as an office and to our clients. There are few architects who have actually sat down and made post occupancy sensor pods that communicate wirelessly to a local and web-based database. When it comes to future installations for our clients we can now speak with the authority that comes with having done it. Since the initial prototype pod LMN has laser cut 8 circuit boards and created more permanent installations placed throughout the office, 2 in each quadrant. Final Pod
In addition to evaluating occupancy comfort, LMN can also test the data generated by the sensor pods against digital simulations from DIVA (daylighting simulation software). LMN has also begun visualizing the data outside of Xively®. This allows us to store the data locally and visualize it however we need. The new sensors have now been up long enough to start studying the data: Improving Data Visualization with Floor Plan Navigation
Some of the initial observations from reviewing the data concern solar radiation and heat gain. Even with the automated shades, we are seeing heat gain during sunny days that is not being offset sufficiently by the HVAC system. To address this issue we now override the shades on sunny days in the west and south west portions of the building to preemptively stop the solar gain from occurring. In the future, we would like to automate this by tying the shades to weather forecasts. We would create a threshold for outside temperature and cloud cover that would override the solar clock just as the shadow sensors do now. The real-time tracking has allowed us to find problems quickly and act to solve them. We are now setting up an email notification system for other systems that require manual adjustment. While the real-time data helps to control the current situation, the data logging allows us to track issues over time and compare the data to find connections between them that we may not have uncovered otherwise. In the end, this endeavor is about designing strategies to make people comfortable in their workplace. From outreach to simulations to subjective monitoring, we are endeavoring to push the boundaries of the field to the benefit of everyone who works at LMN and to those who LMN work with in the future.