As I discussed in previous posts we have a huge, and growing, number of sensors in our cities detecting traffic flows, pollution detectors, safety cameras, smart meters, temperature and wind, power usage and so on deployed and monitored by the Municipality. Additionally we have sensors deployed and monitored by biz and private citizens that a Municipality may “invite” to share. In the figure the “Institutional” sensors, that is the ones deployed and controlled by Municipalities, Institutions and commodities enterprises (power lines, telecommunications, waste management…) are shown on the upper part, the ones that are controlled by single citizens are shown in the lower part (traffic cameras are actually both in the public and private/individual domains). Inductive sensors, like the ones to intercept RFID are by far under the control of individuals, placing in this category shop owners.
All of them are part of the hard awareness infrastructure but they have not been deployed to create “city awareness”, rather they serve vertical silos with very specific goals. It is up to the Municipality to aggregate these silos and create the city awareness.
There is no general solution to this, each silos is aggregating data its own way and may or may be not have been designed for sharing the data.
The existence of a city plan for data aggregation and management would clearly help. Rather than looking for ad hoc solutions a city may provide a general framework for data sharing and request/solicitate the sharing of data from its own silos and from the ones owned by other parties.
Even the sharing of its own data is not an easy walk, it requires effort. Of course, looking ahead, a Municipality can enforce the sharing of data in all new systems under its control, demanding that their design accommodates for data sharing.
There are several technologies and architectures for data aggregation and analyses. The most interesting ones are those that can learn and dynamically adjust to the changing data streams. Deep Learning techniques are today the ones most promising.
The goal is to create a digital signature of the city, continually evolving. Any new correlation is mapped on the existing digital signature and divergence are generating the awareness that something worth considering is happening. At the same time, recurring divergences retune the digital signature in a learning process that keeps pace with the evolution of the city.
Sensors intercepting a high traffic peak at 8 in the morning are actually stating the obvious since every day you have peak traffic in certain parts of the city as people are going to work. On the contrary, a high traffic situation at a time and place where you won’t see happening normally creates an awareness that something is wrong. If you are stuck in your in a traffic jam and you get a message telling you that exactly where you are there is slow traffic … well, this is not providing any awareness (I bet it has happened to you more than once…). on the other hand signalling that a road with normally fluid traffic is now getting congested as you move in that direction is providing awareness.
There is a need for data repositories, both distributed and centralised. Centralisation may be most effective but it may be harder to achieve when dealing with data owned by private parties. In these cases one may accept distributed repositories and access them through queries. This is a better wy to protect ownership and privacy of data. Notice that I am not advocating a physical “centralised storage”. Data may sit in the Cloud(s). I am addressing “centralised” vs “distributed” in terms of ownership and access control.
The crucial point is that a Municipality should design and foster an awareness infrastructure where as many data as possible are accessible and can be correlated. As I said, it is essential to create a digital signature of the city since this is the one that compared with the current situation generates the awareness. An additional goal is to be able to generate “predictive” awareness, that is by interpreting data flowing in and knowing what the city digital signature is and its response to various events be able to provide information on what is likely to happen (there will be a traffic jam in 20 minutes in a specific location). The next step, of course, is to take action to avoid the predicted problem.