Investment managers will always tell you that past performance is no guide to future share prices. But history can always teach us lessons. Let me tell you about the future of the smart city by starting with a little story about the history of the internet.
13 years ago I was working with a company called Exodus. It was worth 37 billion dollars. It had built or bought a good chunk of the internet: the hardware infrastructure that you and I rarely see, fibre optic cables and data centres around the world.
Twelve months later it was bankrupt and being acquired by Cable and Wireless for $800 million.
A lot of capital had been sunk into Exodus and a number of other, similar companies. They too hit the financial buffers and their assets were broken up and sold off cheap. Together the assets these companies built represent a significant fraction of the internet we use today.
Exodus went bust in part because it built its business plan on forecasts from Gartner Group and other analysts about demand for internet services. Like lots of things in the dotcom bubble, these forecasts were inflated. They were accurate in their scale but not in their timing: demand just didn’t scale fast enough to justify the colossal investment.
In fact demand didn’t really take off until about five years later when YouTube and others came along with services that made maximum use of the new broadband access networks. YouTube and other content companies benefited massively from the capital that had been sunk into rolling out infrastructure. Hence why we now have debates about Net Neutrality.
What’s the relevance of all this to smart cities? The people who spend the money to build the platform aren’t necessarily the ones who benefit from that investment in the long term.
A New Internet Bubble
Scroll to 2013 and the new hype is about smart cities. For me a smart city is defined by four characteristics:
- Collect: A sensor network to harvest data on environmental and service variables
- Connect: A communications network to get this data back to a hub
- Process: The ability to turn this data into information and publish it in useful formats
- Act: The capability to react to this data to make meaningful change
The reason there’s lots of hype about smart cities is threefold.
Firstly and most nobly, measuring activity and environmental factors, and responding appropriately could mean cutting energy usage and CO2 emissions.
Secondly and somewhat less nobly, public services are under pressure and there’s a belief that data can help to optimise resource use and reduce costs.
Thirdly, and with very little nobility, the big technology vendors and service providers see a new market opening up.
The enthusiasm of companies like IBM and Telefonica for smart cities has fuelled a lot of scepticism about the digitally-driven smart city concept. But it’s a scepticism that I don’t share.
Today this is a position of faith: nobody is sufficiently advanced in these projects to show hard empirical proof of success. But it makes sense to me that if you can monitor lights left on and turn them off, you can save energy. If you only send a truck to empty bins that are full you will save fuel. If you monitor the humidity of soil and only water grass when it needs it, you will save water. As long as the capital and operating costs are in line, I believe there is a return on investment to be had for the city.
Those costs come down to one particular feature of the smart city. Here’s a breakdown of the major components.
I flew back last night from two days visiting Santander’s smart city project. This is one of Europe’s and probably the world’s most practical smart cities. They haven’t achieved a huge amount yet in terms of ROI, but they are at least doing real stuff on the ground and testing it, in a project led by Telefonica.
What you can see here is one of the standard sensor units used throughout the city to monitor temperature, noise, humidity, air pollution etc. If you’re a geek you might recognise some of the internals. It’s all off the shelf parts that in total cost a few tens of pounds. The most expensive sensor they use in this project is 130 euros and that one is buried in the ground to detect whether a parking space is full.
The whole 12,000 sensors they have deployed to date have cost less than a million euros. Bear in mind that they were developing these sensors as they went and early prototypes cost twice as much as the final models.
This is not a lot of money to invest.
All of the sensors are linked back to base by a network of aggregators and gateways that ship data via various unlicensed wireless standards, fibre that the city owns or 3G. In total across the whole 12,000 sensors the city is only shipping and storing 5MB of data per day.
Once you have the capital expenditure of the sensor network paid for, the operational cost for this is going to be low. And that million euros above? That included most of the communications hardware.
At this point in my tour of Santander I was wondering what Telefonica is doing. There is almost no money to be made on hardware or connectivity, the bread and butter of telcos. So why was this enormous, smart global company so interested in this proof of concept project?
The processing. Once you have all this data flowing in you need a platform to turn it into useful information. This is the hard part, and this is where Telefonica and its competitors see the opportunity for revenue.
Telefonica is developing a cloud-based platform under a European initiative called FI-WARE that enables companies to collect, process and visualise all of the data from their sensors. To make the smart city work you are going to need a platform like this.
The final piece of the puzzle is owned by the city and its contractors. It’s about what you do with the data. This could be straightforward daily decisions about when to collect a bin or turn on street lights. It could be a third party application. It could mean better planning decisions — ultimately the biggest source of potential savings.
So, what is the future of the smart city? Here’s the situation I see.
Financial pressure will drive councils to adopt smart city thinking, and they will buy the vision being peddled by the technology companies. Shared procurement Guidelines will help councils to limit their exposure to risk. They will take investment from Europe and central government where they can but most of the capital will come from enterprise. Big integrators and telcos will fund investment as part of multi-year build and operate deals. Service suppliers will be obliged to incorporate smart features in their tenders for waste collection etc.
These providers won’t suffer like the first investors did in the dotcom crash. But they also won’t be the big winners. As appears to be mandated by the EU development funding, open data will be the norm, with common, open standards for data structures. As long as we can avoid fragmentation, this will support a thriving development community building new apps and services on top of the public data.
Though I accept it is a position of faith today, I believe that making cities smart can deliver both financial and environmental benefits to government and citizens. And I believe that if they are properly executed, they could also provide an exciting new platform for third party innovation and development.
Call me a smart city optimist.