Illuminating Smart Cities

Article By : Rick Merritt

Every city is different in what mix of sensors it wants — or whether it wants them at all.

SAN JOSE, Calif — Cities are upgrading streetlights to LEDs, but they are still on a slow road to the internet of things.

Startup Telensa and Signify — formerly Philips lighting division — have installed a total of nearly 2 million LED streetlights to date, with Signify commanding a slightly larger share. That’s a drop in the bucket of an estimated 360 million streetlights worldwide.

The two companies are just starting to see movement putting IoT sensors into the poles. Telensa recently announced a pilot project using AI, but for Signify, that’s beyond the scope of 2019.

To date, Signify has deployed a little more than 500 streetlights with integrated LTE or Wi-Fi. It plans to launch this fall a model supporting 5G, including millimeter-wave bands and a mix of sensors.

“There’s a lot of work to be done … It can be a challenging sell, but we’re trying be a trusted adviser to the city,” said Bill McShane, national director of Signify’s iCity program.

For its part, a small fraction of the 1.7 million LED lights that Telensa has deployed uses sensors. “We’ve grown by a factor of three in four years, much of that on LED conversions … it’s a wave that’s just beginning … [sensors] offer a benefit, but it’s not an overwhelming business case” said Keith Day, vice president of marketing for Telensa.

Telensa Smart Light
Telensa packs a proprietary 900-MHz radio in a controller pod attached to an LED streetlight. (Image: Telensa)

Connected LED lights are an easy sell compared to IoT. The roughly $60 units can last 25 years and pay for themselves in less than seven, thanks to energy savings and central monitoring features that save truck rolls.

By contrast, working out the business case for IoT and selling it across multiple city departments is a challenge. In many cases, cities lack big data sets and policies to run analytics.

For vendors, the business is “frustratingly different in every country,” Day said. “For example, in the U.S., 65% of streetlights are owned by utilities; in the U.K., they are mainly city-owned.”

Often, one entity owns the streetlights, another maintains them, and a third company made them. “You have to understand the supply chain and its motivations,” he said.

Like most things in government, the projects move slowly. A typical streetlight pilot takes 18 months.

The good news is that “cities want to be smart and utilities want to offer new services, so there’s motivation … there’s a new generation of city leaders coming up, and a lot of pilots related to data services are starting to prove themselves — the U.S. is at the forefront,” he added.

Cambridge is testbed for AI traffic analysis

Every city is different in what mix of sensors it wants — or whether it wants them at all.

“The sensor we hear most call for is air quality, then traffic control, road conditions, noise, and seismic activity,” said McShane of Signify, which is working with a handful of sensor providers for its next-gen streetlight rolling out this fall.

Telensa sees slowly growing demand for a mix of sensors currently led by radar. The sensors can automate traffic surveys for planning and dim streetlights during low-traffic periods. The rise of cheap cameras and radars has largely killed an early IoT market that used magnetic sensors to detect a car’s presence or absence in parking lots.

Today, Telensa’s highest-profile project is a trial using neural networks to analyze video and radar data of traffic in Cambridge, England. It uses Qualcomm’s Snapdragon 845 to track patterns locally that are sent to Microsoft’s Azure cloud for further analysis — all without identifying individuals.

“We discard the video and keep the insights for modeling that goes into the city’s own systems,” said Day, noting that the projects can help cities hammer out data-privacy policies.

Telensa aims to spin up as many as four more trails before the end of the year, two each in the U.K. and the U.S.

The company also works with a handful of third-party sensors, some riding 2G or LTE networks and others on its proprietary 900-MHz network. FarSite Communications provides sensors to check when trash cans are full. Products from InTouch measure when drainage ditches are rising due to clogs.

“Lots of people do air quality sensors, but they have a frequent service interval for changing out filters,” making them more expensive but still a hot topic, Day said.

Overall, streetlight “real estate has potential to be an array of interesting things from 5G cells to air quality sensors — all powered out of harm’s way,” he said.

A look inside Signify’s 5G streetlight

For its part, Signify is upgrading LTE/Wi-Fi poles to house 5G mmWave gear. “We’re taking a mundane piece of infrastructure and making it an asset,” said McShane.

Its first-gen poles had a 32 × 32 × 43-inch base for cellular radios supplied by Ericsson, Nokia, and Samsung. The streetlights coming this fall are relatively svelte and dense. A 14-inch diameter base supports a 7-foot pole packing seven 5-W radios and a mmWave unit at the top.

A separate design packs Wi-Fi base stations from Ruckus Networks and an optional mix of sensors. Keeping thermals under control has been one design challenge for the new units.


Signify’s so-called BrightSites consists of three smart poles. Click to enlarge. (Source: Signify)

Business deals vary. Sometimes, Signify pays reportedly as much as $10,000 to deploy the poles, then rents the space to comms and sensor service providers. “We’re also developing a dashboard for analytics that we will bring out later this year,” McShane said.

Even without IoT sensors, the LED lights still bring a new layer of data services, with each pole handling its own setup, typically via cellular links. “LED lighting is becoming a way to collect data,” McShane said.

Telensa’s LED lights use its ISM-band radios to communicate up to 10 miles at a stately 62 bits/s. The links are a version of an ETSI comms standard, running on a 32-bit microcontroller.

Leave a comment