This is the seventh article in the Precision Construction series, which explores the application of the Internet of Things to digitally transform the construction industry, ultimately with the objective to improve labor productivity, reduce costs and enhance safety. The series began with Exploring Digital Transformation for Construction, followed by Simplifying Complex IoT Solutions, United Rentals Drives Efficiency & Excellence with IoT, United Rentals Helps Customers Optimize Equipment Rental, Robotic Masonry and Mixed Reality for Construction: Applicability and Reality and Digital Transformation – Enabling New Business Models for Construction. Articles generally follow a five-layer framework, described in Simplifying Complex IoT Solutions, that makes it easier to understand digital transformation solutions. To learn more about the various technologies described in this series, visit PrecisionStory.com.

Skanska, a multinational construction company based in Stockholm, Sweden, is the fifth largest construction company in the world according to Construction Global magazine. Notable Skanska projects include London’s 30 St Mary Ax building (commonly known as The Gherkin), University Medical Center in New Orleans, World Trade Center Transportation Hub and the Meadowland Sports Complex—home to the NFL’s New York Jets and Giants and the 2014 Super Bowl.

Many of Skanska’s job sites are environmentally sensitive; there are requirements that vibration, noise or airborne particulates are kept to a minimum during certain periods of time. For instance, the recently completed $27.5M renovation of the Salah Foundation Children’s Hospital was located next to a hospital’s neonatal care unit, so Skanska had to ensure that there was no jack hammering or other noisy activities that would disturb the doctors, nurses or patients at the hospital.

Figure 1: Traditional Jobsite
Environmental Monitoring

Historically, environmental monitoring has been a low-tech, manual process that includes mounted sensors, big red light bulbs and an hourly check by someone walking by each monitoring station to look at sensor readings and update the numbers on a paper log sheet, as shown in Fig. 1. This approach has many obvious limitations, not the least of which is the inability to communicate in real time with Skanska or the customer if there are issues at the job site, such as noise being above acceptable levels. Skanska knew it could do better. 

Things

Skanska’s initial solution, iSite Monitor, was deployed in 2012. It allowed a construction team to monitor a range of environmental conditions such as vibration, noise, air particulates and pressure differential. It replaced pencil and paper with a network-connected device with the ability to provide real-time notification of environmental issues at the job site. In 2014, an updated version called inSite Monitor 2.0 was deployed. This new version didn’t change hardware design, but rather updated the software, which included connecting to Microsoft Azure in the backend.

Figure 2: Hackathon Whiteboard Brainstorming

In 2015, Doug Seven from Microsoft arranged a hackathon involving four engineers from Skanska and a group of engineers from Microsoft. The objective was to create a working proof of concept of a significantly upgraded inSite Monitor in four days, leveraging new hardware under development by Microsoft.

With a lot of duct tape and soldering, they created a prototype that replaced the existing ruggedized, handheld sensors with smaller and less expensive sensors. These sensors included temperature, humidity, differential pressure, vibration and noise. The prototype used MinnowBoard MAX, an open hardware, embedded board that deploys the Intel Atom E38XX series system-on-chip (SOC) at its core, running Windows 10 IoT on SD card. The prototype also connected to the Azure Cloud and deployed Azure Stream Analytics for querying data streams. In addition, data visualization was done through Microsoft Power BI.

Next came the task of getting rid of the duct tape and turning the prototype into a product reliable enough to monitor real world job sites 24/7.

Figure 3: inSite Monitor Modular Design

The next-generation inSite Monitor 3.0 has a modular design, as shown in Fig. 3, enabling individual sensors to be added and removed—depending on the use case requirements—without having to replace the whole device. The sensors currently available to be deployed in the field include temperature, humidity, water leakage, atmospheric pressure, noise, vibration and air particulates.

On the software side it’s no surprise that it runs the Windows 10 IoT Core operating system, which is a very small footprint version of the Windows OS, targeted for small-form-factor devices. Developers may write programs in .NET or JavaScript on their Windows-based computers and then deploy to computers running Windows IoT Core.

Figure 4: inSite Monitor Solution High-Level Architecture

Connect

As shown in Fig. 4, the inSite Monitor also acts as an IoT Field Gateway, connecting sensor data to the backend running in the Microsoft Azure Cloud. Network connectivity to the internet is handled via WiFi, but there are also two USB ports for connecting to a MiFi network if desired. Although each sensor is configured differently, data is typically kept on the device for 30 seconds or so and then average values are sent to the Azure Cloud.

Higher-level communications are handled via the Microsoft Azure Service Bus, which provides the messaging layer between sensors at each job site and the Azure IoT Hub running in the cloud. The Service Bus is a brokered, third-party communication mechanism, ensuring that the data is delivered even if the inSite Monitors and Azure Cloud aren't available at exactly the same time.

Data is transported using the Advanced Message Queueing Protocol (AMQP), which is an open-standard, application-layer protocol for message-oriented middleware. By using AMQP, Skanska was able to build a cross-platform, hybrid application using components built using different languages and frameworks and running on different operating systems.

Collect

In terms of storage requirements, only a few bytes of data are sent per sensor, so over a four-month period, less than 200KB of data will be stored per inSite Monitor in the field.

Once connected, sensor data is collected using Azure Table Storage, which is a NoSQL database for schema less storage data. From there, data is processed by a cloud services worker role and transferred as structured data to an Azure SQL database. This database also stores inSite Monitor configuration data such as names and locations of sensors. Also, when an alarm goes off, a record is generated and also saved to this SQL database.

Azure Blob Storage is available to store user data and image files, although at this time no user data is being stored—only icons for images. Blob Storage is Microsoft's object storage solution for the cloud, which is optimized for storing large amounts of unstructured data.

Learn

As with most early implementations, the insight from the data that’s collected is a combination of visualization and human observation. Skanska uses Microsoft’s Power BI toolset for data visualization. It allows Skanska to see its data organized in different kinds of charts and graphs, which in turn allows them to see trends. For example, Skanska can graph how the temperatures have been trending over the last hour or day or seven days. Skanska can also show various things, such as current, maximum, minimum and average temperatures over the last 24 hours. In the future, Skanska plans to apply machine-learning algorithms to the stored data to search for and detect anomalies.

In order to detect when sensor values go beyond acceptable normal parameters, first an acceptable range for each sensor value is set. Data is then fed from the IoT hub into the Azure Stream Analytics engine, which is a tool for querying data streams as they’re happening in real time. Azure Stream Analytics uses a programming syntax similar to Transact-SQL (T-SQL). T-SQL is a set of programming extensions from Sybase and Microsoft that add several features to the Structured Query Language. With this syntax, Skanska is able to write SELECT and WHERE statements to derive value from the data. For example, Skanska has written queries like “select where temperature is greater than X” or “select where humidity is greater than Y.” In other words, as the data comes in, Skanska has written rules to detect anomalies and warning conditions in the data in real time.

Figure 5: inSite Monitor Mobile App

When a threshold violation is recognized, an alarm is sent through Azure Mobile Services to the Notification Hub and out to the inSite Monitor mobile app. In addition to receiving alerts, the mobile app can retrieve real-time data so graphs can be produced, as shown in Fig. 5. And, given that the mobile app communicates with services running on the Azure Cloud, the app can run on both Skanska employees’ and customers’ mobile devices, ensuring all relevant stakeholders are always kept informed.

Do

With the inSite Monitor’s 24/7 accessibility and alert system, the risk that unsafe environmental conditions at job sites will go undetected has been significantly reduced. Furthermore, sending real-time alerts to all stakeholders allows immediate action to be taken should an environmental issue arise at the job site. Skanska can also demonstrate compliance with both customer and government requirements and regulations.

For instance, all medical construction projects are required to maintain negative pressure to ensure airborne particulates do not drift from the project site to adjacent functioning medical operations. The inSite Monitor allows Skanska to document, validate and respond to any variations in pressure and thereby ensure 24/7 compliance. Finally, trends can be more easily identified should there be systematic issues impacting job site environmental conditions.

As an example, Skanska was contracted to build a five-story addition at San Antonio’s Metropolitan Methodist Hospital (MMH). This renovation included adding two floors of care units along with endoscopy, MRI suites and emergency department support spaces. This 85,500-square-foot, 22-month renovation took place on top of an existing operating facility. This meant that careful planning and communication with the hospital had to be a top priority. In an effort to minimize disruption to the existing operations, the project team incorporated the inSite Monitor. This allowed Skanska and MMH to monitor sensors in real time and receive notifications when a sensor was approaching a pre-set level of risk.

Summary

Environmental monitoring on a construction site is just the tip of the iceberg in terms of addressing the general challenges. Some major Asian cities are already establishing networks of sensors to monitor air quality. New York City Mayor, Bill de Blasio, announced on Earth Day 2018 that the city’s air is the cleanest it’s been since officials began monitoring its quality; and according to the report, air quality has been continuously improving. The mayor’s goal is to achieve the “cleanest air of any large U.S. city by 2030.”

is is an area where people from government, academia and the commercial side are coming together with a particular focus on how best to use low-cost, air-quality sensors to monitor cities, counties and regions. The U.S. Environmental Protection Agency has created an Air Sensor Toolbox to support communities and citizens selecting and using low-cost sensors. Commercial concerns have also emerged, such as Environsuite software for environmental management and Breezometer, which has developed an API for displaying air quality indicators.

In addition to environmental monitoring, buildings are being constructed to be smarter and more instrumented. While monitoring for smoke is important, some startups are emerging to try and improve indoor air quality. One such story begins with Aki Soudunsaari, who experienced headaches from poor indoor air quality in his school in Finland. Fast forward to today where Aki has founded Naava, a startup that builds living walls, putting to use the cleansing effect of microbes in plant roots to improve air quality. Naava removes the plant soil altogether from its walls so the air can interact with the roots directly. To further boost the purification effect, the walls are equipped with fans. Aki’s advisory board includes actor Leonardo DiCaprio and spiritual guru Deepak Chopra. In Finland, Naava already has dozens of major B2B customers from business schools and corporate offices to ice hockey arenas.

Outside air quality, noise and vibration are just part of what will be known in a fully connected construction site. In the future, not only can the environment be instrumented, but also the people, materials and machines.