Pollution and environmental quality may be a global issue, but solutions often come community by community. Community monitoring programs provide ongoing visibility into local air and water quality, alerting officials to developing or worsening problems and enabling them to address those concerns.
In its proposed FY 2022 budget, the Biden Administration seeks $100 million for community air quality monitoring and notification as part of a $1.4 billion environmental justice package.
Sensor data is a key component for such monitoring, and technology advances have made sensors more reliable, accurate and affordable than they once were. As a result, scientists and policy-makers more readily rely on sensor data to understand how human activities can affect climate change. Sensor data can support governments in identifying areas with environmental risks and taking swifter action, especially where vulnerable populations disproportionately feel the impacts.
CGI has supported the collection, analysis and dissemination of environmental data for decades. Today, we help government agencies use sensor data to understand environmental impacts. In late 2016, as part of the Environmental Protection Agency’s Smart City Air Challenge, we teamed with municipal government, academia and non-government organizations to form the Lafayette Engagement and Research Network (LEaRN). LEaRN created a community collaborative to build and deploy approximately 300 air quality sensors within the Lafayette, Louisiana metropolitan area—an initiative which continues to thrive and evolve.
To develop the LEaRN solution, the team engaged our environmental regulatory domain expertise and best practices in data management and hardware design. We also developed a new analytics framework with the ability to scale as needed. This initiative enabled the collection of air sensor data to be shared with the community – real-time environmental monitoring as a public service.
Environmental regulators and policymakers look to real-time data sources to help them better understand environmental impacts. You should consider some of these important design principles:
- Apply an interoperable approach, supported by open standards. Successful sensor-based Internet of Things initiatives require real-time data integration and an enabling analytics platform built on open standards. We use the Open Geospatial Consortium SensorThings API standard—a rich data model that can represent a range of sensor and phenomena types for real-time communication of sensor data. Applying open standards within a scalable and flexible architecture enables real-time environmental monitoring networks to fuse data across domains and data types, such as air quality, hydrology, crop monitoring, flood warning systems and more.
- Push computational load to the edge. IoT solutions enable the gathering of vast amounts of data for processing and visualization. The real-time nature of IoT introduces new levels of network load with associated impacts to project costs. In designing an environmental monitoring solution, it makes sense to push as much processing to the devices closest to the source of the data. Whether applying a push- or pull-based platform, seek to enable the edge device to pack and compress the data, reducing overall network traffic. With LEaRN, we used devices designed to cache raw data and perform initial calculations. We shifted computation overhead from communications monitoring to onboard storage and very restricted curation. This enables the system to function in environments where there is no guarantee of uninterrupted contact with the devices. In general, flexibility around computation load location primes tools to successfully handle low power, rural or remote monitoring data—critical for climate monitoring and sustainable farming initiatives where ruggedized devices must share the burden of data collection and transformation.
- Design for data federation rather than data ingestion. The availability of real-time, low-latency data has grown exponentially, with studies estimating that by 2023, over fifty percent of networked devices will be IoT devices. Operating at scale, sensor-based initiatives require communication across disparate devices and data from multiple sources. When working with real-time data, programs should design solutions to federate that data, rather than ingest it. A digital twin, an important application of this concept, is a modeled version of a physical asset that uses curated subsets of data, rather than ingesting all of the raw data from the source. Representing all available data from a city’s worth of environmental sensors presents a very different task than computing and communicating a series of curated metrics from each sensor. Systems can compile curated metrics at a much lower speed to construct and inform, for example, an air quality index version of the city.
- Take a flexible approach to data publication. Environmental monitoring data serves as a powerful tool to both inform environmental policy and educate the public. Our efforts in Louisiana have become an integral component of the local STEM training, helping young people understand technologies and their environment. Making this data publicly available visually helps drive local action—a key component of many Smart City initiatives. At the same time, academia combines data from multiple sources to integrate it into their research. Detailed, robust data and trend information can help inform public policy at the local level and, when combined with data from other sources, can have implications for policy on a larger scale. Initiatives such as LEaRN aim to serve a wide stakeholder community and benefit from a design that flexibly supports the ability to share and publish data in a variety of meaningful and compelling ways.
The design principles highlighted above directly support organizations and communities seeking to harness the power of data to inform environmental policies, address climate change and address environmental justice issues. With these data, citizens and governments at all levels become empowered to make informed decisions that positively influence both the environment and human health.
LEaRN is an excellent example of just one of the many global initiatives where CGI applies our Kinota™ platform to sensor-based IoT solutions. CGI’s Kinota is an open-source implementation of the OGC SensorThings API that stores data in a scalable cloud-based repository, allowing visualization and analysis for real-time insights.
Learn more about CGI Kinota here.