Could the connected car put the brakes on smart cities?

The term Smart City is vastly over used and lacks a clear definition. It may simply be the rebranding of a street light that can be dimmed, or it could encompass the utopian ideal of an infrastructure comprising street lights, cameras and environmental sensors, using the vast quantities of data captured to improve the daily lives of citizens and reduce the cost of running a city.

Sounds great and a number large corporates such as Philips, GE, Cisco, as well as a multitude of tech start-ups like Sensity Systems, have bought into this dream. The idea is that as cities convert their street lights to energy-saving LED lighting, they trade up to a more expensive version which includes a myriad of sensors and a connected network from which cities can harvest data, provide a potential revenue stream and deliver more efficient services to residents.

The technology can help drivers to find a parking space, avoid congestion and it can contribute in the fight against crime. The problem is no-one is really buying the dream. Yes, you can find trials in cities stretching from Glasgow to Amsterdam to Barcelona, but in the real, austerity-driven world of local government, the vast majority opt for LED; the more exotic may include some off peak dimming, but none have gone for smart. 

Perhaps the biggest issue is that no one has identified the killer app for the connected city – the one that really improves the lives of citizens and the one that brings in revenue for cash-strapped cities."

Why is this?

Established cities have already invested huge amounts in CCTV systems, automatic number plate recognition, environmental monitoring stations and traffic control infrastructure. Most of this was deployed before the strength of the internet had been harnessed. They are bogged down with legacy systems, a vast gamut of protocols, complex contracts and finance initiatives. It is going to be sometime before this technology is refreshed.

Perhaps the biggest issue is that no one has identified the killer app for the connected city – the one that really improves the lives of citizens and the one that brings in revenue for cash-strapped cities.

For a time I bought in to the dream. I had been to the think tanks, conjured up a few potential business models in my head and could convince people in the pub what a smart city could be. However, in the last few weeks I’ve been turned in to a smart city sceptic, or at least the version that is deployed via existing infrastructure such as street lighting. The reason behind my epiphany is a new car.

I’m not a person who changes car very often. My last oil burner expired after eight years and a respectable 160,000 miles. Eight years is a long time in the evolution of a car. The new gas guzzler is half electric and twice as efficient – so much more efficient that the local filling station owner had filled a missing persons report the first month I had it. The biggest difference is the amount of tech wizardry.

The days of the self-driving car are still some time away. The journey to get there is gradual and more and more driving tasks are becoming autonomous, warning you of hazards, taking preventative action before you crash and keeping you informed of what you can expect ahead via sat nav. Cars need a huge amount of sensing technology to perform these tasks. Mine has six separate cameras, a radar speed detector, rain sensor, air quality sensor, plus a whole load of others which I’ll find out about when they start to go wrong.

This kind of technology has been creeping onto cars for years. The biggest difference on this car is it comes with a smart phone app. The app tells me if I have locked the car, where I have left it and how to find it, the stats on my last journey distance, where I’ve been and fuel efficiency. Completing expense claims will be a doddle. While the app and the location technology may sound a bit gimmicky, the stepping stone to the future is that the car is now connected to the internet and is constantly beaming data up to the cloud.

The great news for local governments is that the motorists are paying to deploy the sensor network, and the cost of running it. "

So what does this mean for smart cities?

Let’s take the UK as an example. There are 34.5 million vehicles licensed for use on the road network; that’s nearly seven times as many cars as street lights. If, over the next 10 years, 10% become connected, that is potentially over three million mobile sensors sending data up to the cloud.

So let’s think of the smart city apps which are possible by collecting and processing big data from over three million connected cars. Real-time weather and air quality? Easy. Traffic congestion? Easy. Predicted traffic congestion? Easy, by harvesting the proposed destination and route from sat nav. Accident alerts and location? Easy. Automatic number plate recognition can gather details of the car in front and behind with real time location from onboard cameras.

The great news for local governments is that the motorists are paying to deploy the sensor network, and the cost of running it. Of course, someone will earn revenue from the big data apps, some of which may be passed back to the motorist in exchange for data in the first place. Local government won’t be kept out of the revenue-earning loop, they can still collect the parking fines resulting from me ignoring the where to park your car on busy shopping day app. Actually, they can even buy access to my new iFine app so they know in real time who is parked illegally and where!

Is this the end of the road for lighting in the IoT? No. Loads of other applications in the home and retail are viable and desirable. It is just that automotive technology is running way faster than the world of street lighting and city infrastructure. If intelligent street lights could just be used to deliver public high-speed WiFi, we would be signing up tomorrow.

Pictures:  BMW and Jaguar Land Rover 





Comments 4

Ash, I'm not disagreeing that this technology is coming, but we'll have a lot more abandoned, redundant or legacy systems that are over-integrated in this first rush of enthusiasm than in the future when each system is appropriately composed of integration at the right level. Should a light source be a network device? It should be networked, sure, but should it be the carrier for the network? What's the justification? What's the differentiation? What happens when the network bandwidth needs to be increased. Or security modified? Or connectors changed out? Or more devices added to the environment than they can carry? Do we really want to replace the street lighting of a city? That's insanity defined. But if there's a network and a light and they can talk, great. Then it's down to security, integration and protocols and that's been the brilliance of the recent evolution of the internet of things. The Illumination Devices Internet of Things is still in the DotCom period of "wouldn't it be cool if..." rather than mature design. Cars are a different platform. They have a finite life that can be limited to that of the electronics. For now. So different considerations apply. Plus each vehicle is an individual investment. The LA system that various companies have said they'll automate is an investment in 140,000 luminaires.

It seems to me that the previous comment is right on- the luminaire should simply be a lighting device connected to the communication network, which should physically reside in the pole, and connect to any other devices that may also mount there. The trick is to design a pole that will accommodate this.

I don't agree. Smart City relights will come, it's just a matter of budget and time. Check my piece on this same topic here but thru automotive filters. Same goal, but differenet vertical:-

The issue here is that we are looking at this the wrong way, not having learned the lessons of the internet. With the exception of our multi-function interfaces (pcs and later smartphones), almost nothing on the internet is multifunction. It's low price unifunction. A webcam? A subway arrivals update ticker? An RSS feed of the weather? Lighting manufacturers are holding on to the burning end of the stick. Why would one want to integrate traffic control into a luminaire? What is the merit? I can see using the pole as a physical mounting surface for a traffic device. But if I need to update my traffic device, why do I want to have to update a luminaire too? No, what is needed before we get to the end devices is a platform. Because you're right, there are a lot of legacy systems out there. But they're legacy systems not because the end tools (cameras, traffic lights, whatever) don't do the job that they are supposed to, but because they communicate back via proprietary protocols, often over proprietary/closed wiring to proprietary base stations. They are unsharable. No, if we're serious about the IoT, we need to deal with the protocols and backbone, and once that ecosystem is in place, then we have to work out what features of end devices are complementary and aligned, and which are parallel tasks. If I were a lighting manufacturer, I'd start by building poles with dry enclosures for whatever services may come later, and wireways to make them work, whatever the wiring/power/wireless protocols that come later. Sell a platform and don't pretend to know GPS better than the tech sector. Future proof the physical infrastructure.

Leave your comment