Case Study Summary Findings & Discussion

From OpenCommons
Revision as of 07:26, March 17, 2022 by Pinfold (talk | contribs)
Jump to navigation Jump to search


Wireless
Wireless
Sectors Wireless
Contact David Witkowski
Topics
Authors

David Witkowski.jpegTony-batalla.jpgEssam El-BeikOC.jpgBennyLee.jpgBillPugh.jpgSteve Wimsatt.jpeg

During the development of this Blueprint, the authorship team conducted a series of case study interviews of local government agencies in the United States and Canada known to have deployed IoT networks. The team spoke with the cities of San Diego; San Leandro, CA; Calgary, AB Canada; and the County of San Mateo, CA. Below is a summary of the key findings from these case studies.

Project Structure

The structure of these IoT projects falls into two primary buckets: a) Pilot Projects and b) Production Projects. To illustrate, San Diego and San Leandro both implemented full-scale, citywide production IoT networks initially built for a one use-case, while Calgary and San Mateo County deployed test networks to evaluate pilot solutions across various use-cases. (We define “production deployment” as being fully implemented and used day-to-day government operations. This is contrasted with “pilot” or “test” networks, which have no actual government services using them for operations.)

Pilot Networks

The pilot networks were designed to cover large areas at low costs with no pre-defined solutions in mind. Both projects were developed by the respective Information Technology Departments. These open innovation projects, which involved external partners creating and testing new software on these networks, shared similar goals of iterative learning, exploration, and experimentation.

San Mateo County staff developed a test environment it called “SMC Labs” and collaborated with multiple private sector partners to test many IoT network technologies, including LoRaWAN, NB- IoT, and cellular IoT. County agencies were included in an effort to co-create solutions with private sector firms using IoT sensors and networks. For example, SMC Labs has developed and tested solutions for parking, irrigation, air quality, and more.

Meanwhile, Calgary staff deployed a geographically large LoRaWAN system at a very low cost. It’s IT Department began by working with city departments (“business units”) to identify needs and then worked to develop small proof-of-concept projects around those areas. The IT team minimized risk by starting small and iterating. This was deployed with an agile methodology, setting clear milestones, realistic and working to achieve them. They benefited from an existing partnership between the City of Calgary and the University of Calgary, called the “Urban Alliance,” which brought researchers into the program. 15 This led to a successful “smart agriculture” project at the Devonian Gardens, a City-run botanical garden in their downtown core. Through the process of iteration, experimentation, and learning they eventually developed a low-cost sensor for measuring various plant care metrics.

Production Networks

Cities in our research that built citywide production IoT networks found innovative ways to reduce the normally intense capital costs of constructing a production IoT network. San Diego, CA and San Leandro, CA both employed a similar strategy by leveraging an LED street light retrofit project, expected to save their respective cities millions of dollars in energy costs, to “bundle” the construction work with building an IoT network using the same streetlights as mounting locations for the IoT sensors.

Both cities reported that this approach saved huge amounts of time and money in construction and related costs. San Diego’s project was led by its Sustainability Department, while San Leandro’s project was led by its Public Works Department.

In terms of technologies used, San Leandro deployed “smart” street light technology built on a 6LoWPAN IPv6 network, developed by Paradox Engineering (installed by Climatec, a Bosch subsidiary). San Diego installed IoT sensors developed by AT&T/GE Current, which also use 6LoWPAN over IPv6.

There were some similar lessons learned throughout each project. For example, San Diego identified that it can be worthwhile to have technical staff on-hand during the deployment to test the sensors and tweak them, as needed, and ensure the sensors are working and have proper electrical and communication connections, etc. This can eliminate needing to have crews revisit IoT sensors to perform such tweaks after the installation (thus reducing extra time and cost). Staff also found that using cellular service to connect IoT sensors can be convenient but is also expensive in a large-scale network so, as a result, they are looking for alternative communications methods that do not require monthly service costs. San Diego staff is also engaging with the public over privacy concerns, but this has benefited in having a robust policy drafted that puts these concerns at the fore.

Meanwhile, San Leandro staff discussed the importance of involving technical staff early on to ensure the IoT network design is consistent with a given agency’s current IT infrastructure, competencies, and resources. In their example, the 6LoWPAN network required installing IoT nodes on the street lights, along with some 30+ IoT Wi-Fi gateways located throughout the city, all connecting to the city’s fiber optic network, by way of traffic controller cabinets where the Wi- Fi gateways connected to Ethernet switches. This connected approximately 4,800 streetlights to a virtual server hosted at the city’s data center. As it turned out, this was a major technical undertaking that required several networking components and decisions along the way; as such, having IT involved turned out to be an important factor in the success of the project.

At the same time, the roles of Operational Technology (“OT” – which in San Leandro’s case remained with Public Works) and Information Technology (“IT” – which for San Leandro stayed with the IT Department) should also be discussed early on. For San Leandro, the IoT Wi-Fi gateways became a natural demark where IT transitioned to OT; this provided a framework for support and ongoing maintenance after the network went live. For San Leandro, cross-functional collaboration and informal social networks among staff led to a good working relationship between IT and OT. However, in the absence of strong inter-departmental relationships and trust among teams, operational boundaries should be formally discussed as early in the project as possible.

Conclusion

IoT technologies are still emerging and extremely new ground for local governments. For those who decide to forge ahead and build their own, our case studies outline two distinct paths: a) low cost pilot networks that evaluate applications in partnership with private sector firms; and b) production networks with a single focus such as street lights that can be bundled into larger construction projects that have organizational and political momentum and backing.

For pilot networks, it is important to start small, build the right partnerships (both internally and externally), set the right expectations among stakeholders and learn from each trial. Meanwhile, for production networks, there are huge economies of scale in bundling IoT with other construction projects. However, it is critical to have the right teams involved and develop a “big picture” strategy for how the IoT network can be leveraged for use-cases in the future. Within each path, it is important to engage the public and address their concerns as part of the deployment and policy process. Editor’s Note: Full case study reports can be found in the Full Case Study Reports