Guiding Lights


Street lighting offers a go-to platform for interconnected smart cities; and dual-mode IoT connectivity is smoothing the transition

The dream of an interconnected smart city, where digital technologies knit together multiple distinct civic functions to bring efficiency and intelligence to operations, is a heady one. It proposes that city-living, which will draw 70 percent of the global population by 2050, can be healthy, happy and safe. Crucially, it promises it can be green - that humankind has one last card up its sleeve to stop short of wrecking the planet.

But smart cities are hard. New technology is expensive, local governments are hard-up and politics turns on short-term electoral cycles. A model for centralized technology deployments, which delivers operational and financial efficiencies, and repeats across urban locales, either globally or nationally, is out of grasp. In truth most of the leading smart cities in top-ten lists in glossy news titles are only really a collection of disparate tech pilots and district side-projects. Nothing looks like it can scale.

Think about garbage bins and parking lots, made smart with sensors, and run-through with analytics; ROI in such cases is difficult to calculate and standardize, especially when government is so fragmented – between public authorities and private services, and between towns, cities, regions and countries. Think about air quality monitoring; how does a city easily calculate the impact of clean air on health services? The logic says smart cities are hard, so do the sceptics.

But there is a light on the murky horizon of digital change. Street lighting, of all municipal services, affords a platform for cities to get smart and unite multiple applications for the first time. Consider the varied smart street-lighting projects in San Diego in the U.S. and Copenhagen in Denmark, among an increasing number. These combine an array of sensors in modular hardware units affixed to light poles, enabling remote control of the luminaires themselves, plus an engine to run other functions, such as traffic counters, air quality monitors and even gun-shot detectors.

From up high, on light poles, cities have started to get a handle on the ‘liveability’ of their streets – of traffic flow and mobility, noise and air pollution, and of new commercial opportunities. Even parking sensors, traditionally buried in tarmac, can be hooked up cheaply and effectively to lighting infrastructure. Whole cities can be networked and optimized, suddenly, without digging streets or renting space, or resolving abstract calculations about healthier living and safer streets.

It promises to work because the calculation in most smart street-lighting cases is not gambled in the first instance on the savings delivered by the intelligence of the solution. Instead, the feasibility of digital revolution in cities is a serendipitous consequence of concurrent advances in lighting apparatus.

The energy savings afforded by the switch from incandescent sodium bulbs to solid-state LED lighting, along with the ready supply of power and the ubiquity of lighting infrastructure, have made smart cities viable.

The pace of LED swap-outs is flat-out already; the trend for smart lighting has a head of steam up too. Nine out of 10 streetlights—out of 363 million globally—will use LEDs by 2027, says smart infrastructure analyst Northeast Group; a third will be running smart apps as well - from a standing start a couple of years ago. Until the sums are worked through and blueprints disseminated, nothing else makes the case for smart cities at scale like street lighting, as a networked infrastructure from which to hang all manner of digital pyrotechnics.

Piggybacking on LED savings

The rule-of-thumb, presented by lighting and sensor makers, says smart lighting delivers a 50-70 percent reduction in the management and maintenance costs associated with the infrastructure. But most of that – about 50 percent, enough to swing the case – is achieved just by switching to power-efficient LED bulbs. The rest comes from connecting and controlling the luminaires and delivering sparks of intelligence about their working status across the lighting network

Maintenance costs can be slashed dramatically just with centralized tweaks and insights. The ways are multiple and all add up: Scheduling, seasonal controls and trim times; fault diagnosis and reduced truck rolls. The impact rises with the scale of the lighting network and flows back into the original ROI case. This way the outlay can be clawed back in about five years, says the market, and potentially less with the integration of ‘softer’ smart-city concepts – like those parking sensors, traffic monitors, air quality controls and gun-shot detectors.

Analyst firm, Guidehouse Insights, tracks 200-odd cities to gauge the pace of urban change; a quarter are rolling out smart lighting initiatives, it says. Sales of smart systems are skyrocketing; ABI Research calculates global revenues will jump ten-fold by 2026, to $1.7 billion. The planet’s ‘lightbulb moment’ is just that; street lighting infrastructure, mapped to human activity, is the way forward as a platform for smart cities in a wider context. Over two-thirds of new street light installations will be tied in with central management platforms as early as next year, says ABI, to integrate data from multiple smart city sensors.

“There are additional opportunities to be had by smart city suppliers leveraging street pole infrastructure by hosting wireless connectivity, environmental sensors and even intelligent cameras,” says Adarsh Krishnan, principal analyst at ABI Research. “The challenge is finding a feasible business model that encourages deployment of multi- sensor solutions cost-effectively at scale.”

The question is no longer whether to connect, but how to connect – and how much to connect from the start. This is partly about business models, as Krishnan observes, but funding has started to flow in the smart city space via cooperative public-private partnership (PPP) vehicles, where the financial risk is shouldered by the private enterprise in return for a stake in the venture’s success. Subscription-based ‘as-a-service’ contracts, spreading investments into the payback period, have also stimulated activity.

But the smart city question is also about the choice of technology. Different applications make different demands in terms of coverage, throughput and security. Street lighting has developed around different technologies in different markets. In the U.S., Zigbee-based short range mesh networks are a mainstay of smart metering with an 80 percent share, according to ABI; the same utilities own most of the public lighting network as well and Zigbee’s utility-focused SEP (Smart Energy Profile) 2.0 profile is being reworked for lighting.

In Europe, by contrast, street lighting is being connected with traditional cellular (2G through LTE (4G), commonly), as well as the new cellular IoT LTE-M standard. Proprietary ultra-narrowband (UNB) technologies are also in play, along with Zigbee, smatterings of Bluetooth LE and IEEE 802.15.4 spinoffs.

The Bluetooth Special Interest Group (SIG) in particular is serious about smart cities. It predicts five-times growth of Bluetooth LE shipments in the space in the next five years (to 230 million annually). Most is linked to asset tracking in public venues, for example airports, stadiums, hospitals, malls and museums. But Bluetooth LE is pitched for outdoor networking too. “Asset management solutions increase utilization of smart city resources to help lower operational costs,” says the Bluetooth SIG.

Two technologies are better

There are arguments for each, although certain of them unravel under closer interrogation. UNB, for example, imposes tighter limits on payloads and delivery schedules, precluding parallel support for multiple sensor applications or for needier ones like cameras. Short-range technologies are cheaper and offer greater throughput for developing ‘lighting as-a-platform’ setups. Importantly they also bring redundancy in case the WAN signal drops out and a means for technicians to get direct access to sensors for commissioning and diagnostics. Bluetooth LE, for example, is interoperable with almost all smartphones on the market.

But while a denser mesh creates robustness, it also brings architectural complexity and places higher energy demand on interconnected point-to-point sensors. There is an issue with range too; coverage tops-out at a couple of hundred meters with Zigbee and Bluetooth LE. And while short-range technologies are clearly contenders and well-suited for meshing neighborhood-wide sensors, they are closed networks that ultimately require a gateway to get the signal back to the Cloud.

In the end, a cellular connection is usually integrated into the mix. The trend among smart lighting providers is to go with point-to-Cloud cellular, offering coverage of five-to-15 kilometers from gateways or sensor devices. Cellular brings range and simplicity; it also provides ready- built networks and higher-grade security, according to the cellular community. “Mobile operators ... possess total coverage over urban areas, [so] no additional infrastructure is necessary to connect city lights and sensors,” says Neill Young, IoT Verticals Lead at the GSMA, the industry organization that represents the interests of mobile network operators. “The security and reliability of ... [cellular] networks in licensed spectrum [means] operators are best-placed to support large numbers of low cost devices requiring long battery lives, minimal maintenance and long ranges,” adds Young.

ABI says cellular, out of all the connectivity technologies in play, will see most growth in the next few years. The clamour for 5G networks, and the struggle to host 5G infrastructure, has seen operators seize on light poles for small-cell infill in urban environments. In the U.S., Las Vegas and Sacramento are rolling out LTE and 5G, plus smart city sensors, on street lights with carriers AT&T and Verizon. And Hong Kong has just unveiled a plan to install 400 5G-ready lamp posts as part of a smart city drive.

But LTE and 5G are geared for more rarefied smart-city cases, requiring higher throughput and lower latency, such as high-definition camera surveillance. The IoT sector at large is propped up by low-power systems in the form of short-range technologies like Bluetooth LE and Zigbee, and cellular IoT LPWAN technologies NB-IoT and LTE-M. These can be combined to powerful effect in street lighting platforms, and elsewhere in smart cities, as part of lower-rate sensors bringing intelligence about city functions and environmental conditions.

“The solution is to team inexpensive but range-and resource-limited short-range wireless with cellular IoT so they complement each other. That way the data—about traffic flow and footfall, air quality and temperature, vibration and noise, or whatever else—can go from the sensor network to the Cloud via a secure and robust cellular network,” says Svein-Egil Nielsen, CTO with Nordic Semiconductor, a short range wireless and low power cellular IoT solution provider.

Tight integration in hardware

Nordic provides multimode short- and long-range products. Its nRF52840 SoC supports Bluetooth LE, Bluetooth mesh and Zigbee, as well as Thread and proprietary 2.4GHz systems. Nordic’s cellular-based nRF9160 SiP offers both LTE-M and NB-IoT. “The combination of the two technologies brings advantages in terms of performance and cost,” adds Nielsen.

Frequency separation allows these systems to coexist, with the former running in the unlicensed 2.4 GHz band and the latter going wherever LTE goes. There is a trade- off between wider-area coverage and larger-capacity throughput at lower and higher frequencies. But in lighting platforms, short-range wireless is being commonly deployed to interconnect sensors, edge-based compute power is being charged to direct insights and cellular IoT is being used for backhaul to the Cloud, and for higher- maintenance sensor controls.

As yet, twin short and long-range radios are being added separately; they are not embedded into the same silicon in the factory, and there is a case on one hand to keep components apart, as luminaires, sensors and radios all fail differently. At the same time, embedding twin radios into a single system will bring tighter technological integration and lower acquisition costs – which are prime considerations for smart cities.

The market is going that way, says Nordic; the firm has already combined short range wireless and cellular IoT connectivity technologies in hardware and software, at the developer level, for solution makers to run the pair together in test applications. The company’s single-board DK for the nRF9160 SiP is designed for developers “to get their cellular IoT applications working”; the Nordic Thingy:91 is described as a “full-blown ready-to-go gateway,” available as an off-the-shelf prototyping platform or proof of concept for early product designs.

Both feature the multimode cellular nRF9160 SiP and multiprotocol short range nRF52840 SoC. Embedded systems combining both sorts of technologies for commercial IoT deployments are only “a few months” from commercialization, suggests Nordic.

“Smart-city lighting platforms are playing host to all of these connectivity technologies already; the market shows very clearly the demand to combine them, and we’re already providing developer boards for solution makers to test how they work together. It is only a matter of time before they are combined in commercial solutions, as well,” says Nordic’s Nielsen.


Smart cities are hard; new technology is expensive, local governments are hard-up and politics turns on short-term electoral cycles