“Smart cities” comprise increasingly dense networks of remote sensors, providing levels of insight never before achievable and made even more valuable by allowing the correlation of diverse sources. The promise of smart cities can only be unlocked with real-time data access, yet many urban sensor applications prohibit wired connectivity and cannot be supported by short-range WiFi. The use of industrial cellular gateways eliminates the barrier between data and cloud databases in systems where measurements would otherwise be stored only on disconnected local storage. Cellular gateways retrofit to existing sensor deployments, and as such, a gateway can be used in lieu of installing an entirely new connected sensor system. Further, specialized gateways reserve flexibility for the end user to adapt to certain design decisions such as the use of independent power versus wired power, and the type of battery to be used in accordance with the deployment environment. For functionality, cost-effectiveness, ease of integration, and security, cellular gateways are a sound choice to implement and complement scalable SCADA systems used for remote monitoring.
SCADA and Cellular Gateways
Automating the acquisition of data is critical, especially in the modern world of aging infrastructure. That’s where supervisory control and data acquisition (SCADA) comes into play. Systems administrators can reduce costs and increase insight regarding urban metrics, such as the integrity of structures such as roadways and bridges, by using well-designed SCADA systems.
Properly selecting hardware is a crucial process in the SCADA development cycle, affecting later business decisions such as ROI time and scalability. While off-the-shelf programmable logic controllers (PLCs) with connectivity modules work for many industrial applications, they can be insufficient for exceptional use cases. Applications that are demanding on hardware, such as distributed agriculture monitoring, marine sensing, and urban networking, are much better served by packages known in the IoT industry as cellular gateways.
There are currently two approaches to connected systems design, both having serious flaws in sustainability. The first approach is to rely on custom engineering as the solution to any and all hardware needs. Some deployment scenarios demand this research and development, but in most scenarios, contracting custom engineering solutions is akin to reinventing the wheel. A PLC/connectivity solution can be built from scratch, and it will get the job done, but it is grossly wasteful in terms of time and financial investment. In the worst case, if a municipality is designing a connected system and contracts a custom engineering firm for a solution, they might find they are subject to a restrictive exclusive technical relationship, thus barring them from integrating with suitable existing products in the future. While the work done by engineering firms is undoubtedly valuable, designers should first look to work with IoT product manufacturers to source proven solutions at a lower cost, and have custom solutions developed only when existing solutions are incompatible with client needs. The customer will benefit by having a stronger device that is easily maintained.
The second misguided approach to systems design is the polar opposite; in cases where custom solutions are not developed, designers turn to siloed products with little adaptability. Silo hardware providers are well established and create rigid hardware products that do one task very well, but are designed (in terms of geometry, application layer logic, power requirements, etc.) to perform best in the situation it was designed for, and only in that specific situation. These systems might not have been designed for your current application, might not contain device health self-reporting capabilities, might not be able to meet complex mixed-material RF environments, and perhaps most important, might not effectively balance price points to make scaling the solution economically feasible. A device that measures only groundwater levels is not as valuable as a device that measures groundwater levels today and can also measure volatile organic compounds (VOCs) next week and a new metric the next through simple sensor swaps. Large-scale deployments thrive, in functional and economic terms, when built around reuse and modularity. Siloed networks and siloed systems are less valuable networks and less valuable systems than their adaptable counterparts.
When considering design choices, take a step back, look at your long term goals of your connected system, and weigh the cost of having an Engineering Development Model constructed and having a design transfer performed versus modifying an off-the-shelf solution. Give extra consideration to existing solutions that are designed for adaptability and customization—while you’re best off not working at the circuit board level, you also don’t want to invest in what are essentially plastic bricks that are completely inflexible. Proven solutions that have a durable form factor but also allow for reconfiguration have an immense long-term value. The hidden consideration in solution comparison is providing for intuitive device maintenance and servicing. Maintainers will likely be parties other than the manufacturer or the owner of the device—and if you think servicing can be done quickly and easily on a poorly integrated piecemeal system, you haven’t thought deeply enough.
Analyzing the Needs of a Smart City
Smart cities are complex feats of engineering, requiring good design to be safe and valuable. Good design stems from experience and is reflected in a uniform and tested product, proven to work in several sectors while maintaining best practices for security, device health, and user safety. The ability to maintain best practices, and ensure the longevity of the device, is a direct benefit of purchasing an industry-proven IoT product as opposed to having a custom one-off solution made. A product will have a support and integration team, whereas a custom solution typically loses that support rather quickly after the contract is complete; this is why dependability problems have risen with custom engineering firms in the smart city space working without a product provider partnership.
Resiliency and cost reduction are often said to be two of the most important traits of sustainable IoT, and they are best ensured by proper hardware and network selection. Network selection remains more challenging than hardware selection, but the use of cellular gateways provides for additional flexibility in network implementation. In many applications, complex RF environments require custom-modified solutions to improve RSSI in mixed-material urban environments; in some cases, embedding cellular transmitters in the soil or in monitoring wells beneath sidewalks offer the only viable connectivity solution able to provide access to the desired data and overcome the limitations of infrastructure surrounding the sensing system.
Industrial off-the-shelf SCADA solutions sometimes don’t have the proper specifications to function and survive in smart city environments, whether that is due to issues in connectivity, durability, or scalability (pricing, maintenance time, and expandability). Using a well-designed cellular gateway hardware and software framework as starting point for your solution allows for a great amount of customization in order to overcome challenges and problems that can’t be foreseen before deployment:
Data streaming can be modified. This allows for easy adjustment of simple dynamic settings such as time between samples, more complex behavioral changes redefined by updating the firmware of the unit to support new complex features, or even rollback capabilities in the event of a budget cut.
Additional hardware can be retrofitted. With easy remote updates, this opens an area for city workers or third party contractors to easily go out and modify the devices. When developing smart city infrastructure, maintenance is key, just like it is for infrastructure in general. Any smart city solution, cellular or otherwise, needs an easy-to-upgrade system for sensors. Ultra low-cost sensors often need to be replaced entirely, as they lack the ability to be effectively calibrated after a relatively short expiration date. Even higher-end sensors often need to be replaced as the equipment required for field calibration is rather unwieldy. Therefore, any solution needs to be able to have its firmware updated to support changes to sensor specifications from third party vendors.
Distant devices can be coordinated. Mesh networks have the benefit of not requiring payment to a third party network provider, and are thus a common solution. Local area connections such as these can suffer from interference if badly designed, and many others draw a surprising amount of energy and so the benefit of not paying for a network is offset by the operational and business expenses incurred by operating an isolated mesh. But, they lack some of the great remote diagnostic and security capabilities that a cellular network provides. Mesh networks coupled with farther reaching networks such as cellular or satellite networks are a great hybrid approach that give you the costs savings of mesh with the reliability of the cellular system. Frequently, cellular gateways are used as ports to the Internet for large swaths of local area sensors that connect to the gateway via a local mesh strategy. Coordinating devices securely on the implemented network enables correlation of localized data for unparalleled insights, and choosing a cellular network enables remote access and coordination.
And, by the top-down perspective, cellular gateways can easily be retrofitted to existing sensor systems to make them remotely accessible and controllable, making a city smart with minimal redesign and reengineering.
Process of Integrating a Cellular Gateway
Integration is one of the key drivers of design. The ease of updating basic hardware necessities, like power packs, is key in urban environments given the scale of active sensor systems and the uptime that city administrators expect for their investment.
Designing and integrating IoT systems, whether custom made, off-the-shelf, or a hybrid solution, starts with determining power management, and determining power management starts with a detailed understanding of the application. Simple alert-based systems have long idle times (at low power consumption) waiting for interrupts; however, other systems require server-side synchronization and messaging by the client (at varying levels of power consumption). Clearly defining user requirements and analyzing tradeoffs is key for IoT systems, especially in a world where customers are expecting more and more from new products. Customers must understand fundamental limits created through power consumption, the balance of network usage, and precision of timing. Figuring out requirements for the precision of check-in times, balanced with the cost of the device, the use of the network, and power usage, is imperative. IoT and smart city systems are most successful when the end user provides design restrictions so the perfect device for the task at hand can be determined.
Once logistics and expectations are set, physically integrating cellular gateways in smart city deployments is fairly straightforward, especially if using an integrated stack or services built for IoT deployments: In the age of SaaS, connecting a user interface to a datastore and then having a secure middleware to route connections to and from the device is an easy task. If data streaming behavior is preconfigured by the gateway manufacturer on your behalf, integration is as simple as unpackaging the gateway, hooking up your sensors, and connecting the power supply (which might be external wired power or an internal battery pack). Depending on the manufacturer, data will start streaming immediately, or some final activation steps such as configuration of routing and datastore read/write permissions might be required. From here on out, configuration and reconfiguration can be done remotely through software.
Securing Cellular Gateway Systems
Security should always be one of the first items on an integrator’s mind. Cellular-based systems have some benefits to providing inherent network security, but at the same time have weaknesses that require attention just as in any system. Here are some key security requirements for cellular gateway solutions:
OTA Updating. Out of the box, many cellular packages support verified OTA updates delivered through network tunnels, allowing for remote updates to install patches issued by the vendor or by the firmware provider of third-party hardware components.
“Device Spoofing”-Proofing. Well-designed systems work using key authentication on the client and network side to verify identity. In this way, the device and datastore pair reduces the risk of device spoofing. Cellular devices lend themselves to secure applications since cellular IDs are globally unique and allow for easy administration, but need some help from the user with regards to authentication. For messages from or to the device, a strong authentication scheme is required beyond using only the device UID, since SMS can be spoofed to give incorrect UIDs.
Remote Shutdown. In the event that a device is compromised and begins sending unwanted messages, SIM cards can be remotely deauthorized from accessing the cellular network. Of course, this does not prevent the hijacked device from remotely controlling local assets, such as water pumps in a groundwater monitoring system, if the attacker has a network attack vector beyond the cellular network. But, direct control over each cellular connection, thanks to the economics of maintaining cellular networks and providing service only to approved devices, allows administrators to effectively “brick” a cellular device if it is hijacked.
Encryption. There have been attacks against the GSM encryption standards, such as A5/1 decryption attacks directly used by the NSA (as shown in documents leaked by Edward Snowden) and protocol reduction attacks which operate by downgrading to a different A5 standard, such as type 0, which uses no encryption, by using a spoofed base station. This being said, A5 security is always better than no security, and this level is built right into the cellular networks. Still, care must be taken to prevent reduction-in-standard attacks. Decryption attacks require equipment and computational power, and for most applications, A5 will be sufficient. Putting more computational power on the edge to perform additional encryption defined by another stronger protocol is a valid workaround for particularly sensitive data, but for low power sensing and alerting applications where nano-amp draws and multiyear battery lives are part of the minimum viable deliverable or POC, the costs of integrating the necessary power to handle stronger encryption must be considered.
Viral Spread Mitigation. Having control of a single cellular device does not compromise your entire system. Some other network technologies allow for bad application design to create situations in which peered devices can control and interact with each other. Cellular devices have this weakness as well, but such connections must pass through the application server, which has the resources to inspect and analyze interactions for foul play, thus lessening the risk of a total network exploit resulting from poor security design. The critical point of cellular networks is the application server, and care must be taken to ensure control of it. As an extreme worst case backup, maintaining direct control of the billing system that pays for the device’s service means you always retain the ability to shut down devices if needed to stop viral spread.
Moving Towards a Smart Future
In a world of barriers between sensors and stakeholders, cellular gateways bridge the divide to allow data access from where it is most valuable and actionable. This issue is easily conceptualized for distant marine sensing deployments or widely distributed agricultural monitoring systems, but the same considerations can—and should be—applied to the emerging smart city space. The demand for data is increasing exponentially, and systems that are being designed now must be made sustainable for the coming years. A solution that is not scalable is one that will have to be replaced entirely. But, a solution designed with scalability as the foremost consideration will pay dividends indefinitely, and can open new frontiers in connectivity with levels of insight never before seen. This solution takes the form of a cellular gateway, and it’s not going away.
Daniel McCormack is the CEO and Director of Engineering of BluCloud, Inc., an IoT hardware company specializing in rugged cellular gateways which also offers the BluBase data analytics cloud platform. Dan has an extensive background in electrical and software engineering, was the lead architect of the BluBase platform, and designed BluCloud's line of cellular gateways. Dan is also available for engineering consulting services to tailor BluCloud technology to your deployment. Contact: email@example.com Learn more about the BluCloud IoT stack, from sensors to networks to analytics to the cloud, at www.blucloud.tech and follow BluCloud on Twitter at @blucloudtech.
Michael Morscher is the Director of Operations of BluCloud, Inc., an IoT hardware company specializing in rugged cellular gateways which also offers the BluBase data analytics cloud platform. Michael has a background in software engineering and technical management, and was instrumental in expanding BluCloud's hardware line from sensor packages to cellular gateways in response to market research and client demands. He also provides front and back end development services for BluCloud's software integrations and coordinates custom consulting. Contact: firstname.lastname@example.org Learn more about the BluCloud IoT stack, from sensors to networks to analytics to the cloud, at www.blucloud.tech and follow BluCloud on Twitter at @blucloudtech.