Hortonworks’ VP International on the big data challenges of smart cities
Traffic jams, congested roads and bottlenecks, in both the supply and disposal of goods, are ever-present issues and of growing concern for modern cities. The big question for city operators is how to address these challenges and keep their cities flowing smoothly. Is the smart city the answer?
Often, those responsible for digital transformation see data not only as a solution for existing problems, but also as a means of facilitating future development.
As city IT systems advance, the basic tasks within a municipality—such as rubbish collection, street lighting and tourist information—are gradually becoming automated and the once futuristic concept of a smart city starts to evolve into reality.
The Need for Speed
The transport sector is already experiencing relatively rapid digitalisation and so is likely to be an early beneficiary of digital transformation.
Smart management systems for carparks, already being implemented widely across some parts of Europe, use sensor data to query the availability of parking spaces and report back to the municipalities’ central servers.
Solutions such as this are incredibly useful and enjoy a high profile with the public, but the consumer data they rely on can only be captured from specific locations — in the carpark for instance.
Although the datasets from these type of applications are not particularly meaningful in isolation beyond their official use, a more complete picture emerges from an overall analysis of all the data collected from every data-capture system located around a city.
Big Data frameworks are available to crunch vast quantity of information in data lakes, preparing the data for rapid analysis in real-time. Speed is important because the older the data, the less it can be used to assist in real events that are happening now.
An example is data that can help manage and control traffic flows efficiently—a crucial goal for all local authorities, who need to reduce congestion and the environmental pollution caused by unnecessary emissions.
A city and its citizens can also benefit from data provided by private smart navigation systems, which continuously monitor the roads and can transmit their data to a municipal traffic control centre, helping to provide real-time information on accidents, traffic jams or rescue services’ missions.
Similarly, sharing individual vehicle status data—tyre pressure, age of headlight bulbs, date of MOT etc.—from a municipal fleet including fire engines, ambulances or police cars, can assist in lowering public sector costs through improved management.
Data Intersection Points
The energy sector with its smart meters, is another data source that can be tapped for smart urban planning.
Local authorities’ energy consumption rates, as well as those from businesses and private households, can be captured to enable better resource management.
But just like car park data, which has limited value when taken in isolation, data from one industry sector shouldn’t be siloed and there is a potentially valuable point of intersection between energy data applications and the transport sector.
For example, in the emerging area of electric mobility, charging stations located throughout the city would make electric vehicles more attractive but, in order to provide the appropriate amount of energy as and when required, traffic data should be linked to data from the electric vehicle itself and also the power grid so that a meaningful picture could be created.
The fields of transport and energy help to illustrate what a smart city of the future might look like. Since data acquisition is decentralised across multiple distributed sensors, new networks such as the Internet of Things (IoT) will need to come into play.
An IoT network and all the associated legacy networks and connected devices that might have been in the field for many years, is an extremely granular infrastructure that brings its own challenges and requires a completely new approach on the part of the municipal data centres.
Legacy Solutions Leave Gaps
Until recently, the first-choice solution would have been a classic enterprise resource planning system (ERP) but these are not sufficient for processing today’s Big Data applications.
ERPs have a few main problems here: they are not able to record and process such a high density of data in real time, they often lead to silos therefore no way to get a holistic view of all data within an enterprise, the data they collect is designed to produce results for, and be used by, a corporate resource planning system not a whole municipality.
If ERPs are not an option, then a thorough assessment of dedicated Big Data applications is what is required.
First and foremost, local institutions must be certain that the security of the data collected is assured, so a platform that performs automatic data security checks, issues security risk alerts and is capable of real time issue resolution should be top of the list.
It is then advisable to choose a solution capable of eliminating unnecessary costs and silos.
For example, if each employee group within a municipality works in its own separate data silos, there is a danger of the same data being processed more than once and generating additional cost.
A solution which cannot combine all its various information sources not only runs the risk of redundant data processing but is also likely to be inflexible and unable to scale.
Eliminating silos also ensures that users can run analytics using all data, rather than only analysing a portion of the data which may lead to an incomplete picture or result.
Smart Cities Need a Modern Data Architecture
In the creation of a smart city, data from various highly distributed sensors, as well as from other sources such as office documents, must be collected, protected, combined, correlated and processed in real time. Organisations must adopt a modern data architecture to adapt to the new paradigm data being generated in troves. This is no small task and city operators should not shy away from the necessity of a purpose-built open source Big Data platform, capable of cost effectively processing petabyte-level data volumes whilst keeping that data protected.
Open source provides scale with very reasonable economics, and also reduces vendor lock in, flexibility and rapid innovation…
Furthermore, because cities are dynamic, changes in data demand must be anticipated so it’s important to invest in a platform with the flexibility to meet changes in data demand through the smooth addition of extra calculation nodes to the server pool.
The choice of Big Data platform is paramount, and as cities rely more and more on digitisation, the efficient use and protection of data becomes vital, so much so, that a Big Data capability is integral to the creation of a smart city.