Waste and traffic management apps come from internet of things

For years, the fridge has been the poster boy for the internet of things, but lack of commercial availability raises questions about the future

For years, the fridge has been the poster boy for the internet of things. Suppliers imagine the connected domestic appliance anticipating 'off' milk and avoiding shortages of cheese, but lack of commercial availability raises questions about whether a multitude of online sensors might be better applied to things we discard.

Refuse collection in the Netherlands is already benefiting from the IoT. Dutch waste management firm Rova analyses thousands of datasets from multiple sensors and information sources, including GPS, smart devices and RFID tags. These provide streams on truck location, traffic congestion and bin volumes to help them optimise truck routes and bin collection times. This strips inefficiency from the waste management process and has created a 20% operational efficiency saving, the firm says (see below).

Gartner has forecast that by 2020, 30 billion devices will be connected to the internet, adding $1.9trn (£1.2 trn) in value to the global economy. The signs are that early applications are more likely to be found in industry and infrastructure, rather than consumers’ homes.

In the UK, BT is working with Milton Keynes Council on the MK:Smart project, which is connecting devices detecting parking availability and traffic movement to help the town cope with an expected 60% increase in traffic (see below).

Industrial applications

Meanwhile, industrial applications of the internet of things are gathering pace. Since 2012, manufacturing giant Siemens has been involved in Industrie 4.0, a joint initiative between German industry and government, in part designed to exploit IoT technologies.

Paul Hingley, Siemens' data service business manager, UK and Ireland, says the manufacturer is already employing IoT technology in its Amberg factory. “The crux is turning data into real information, being able to use that to form intelligence,” he says. “We are embedding this into our production processes, taking data from shop floor and using it to improve production and maintenance effectiveness.”

But he says manufacturers employing the IoT in their factories should be ready for its impact on their approach to data management and data infrastructure.

“There will be a massive explosion of data: that is the fallout,” he says. “Siemens and our [industrial] customers have to understand this mass of data. We have developed, on a global basis, operational centres to allow for data collection and remote access to it through dashboards hosted by Siemens. Customers can have continuous data stream from embedded devices. This is a brand new business for Siemens, build from the ground up.”

More about the IoT and business applications

Hingley adds: “We don’t know how much data is going to be generated. You can only estimate, and there is no way of validating those estimates.”

Meanwhile, the industry requires technology suppliers and IT departments to cross boundaries that have grown up between business and industrial computing.

Demands for reliability, security and safety mean embedded industrial systems work to different standard and are managed by different people with a different culture, compared with business systems such as ERP, says Hingley. And yet the data from industrial systems will benefit the wider business.

“Operational technology is quite conservative,” he says. “That is where we have the demarcation line between it and business technology. There will be more interfaces that are easier to use in both worlds, so barriers will come down. We feel the cloud will play a bigger role as a method for collecting data. That will be the big leap.”

Understanding the impact

Gartner vice-president and analyst Ted Friedman says businesses are at the early stages of understanding the impact of the IoT on data management and infrastructure.

“When you ask industrial manufacturing companies what they do with the data from all the sensors they have fitted, after its initial purpose, they say they don’t know,” he says.

Business have yet to understand how to store and exploit the data resulting from the IoT, or how it will fit within existing data governance models.

Friedman says businesses need to consider the implications of IoT data across their infrastructure and management processes, deciding whether to adapt existing approaches and technology or build fresh resources to support IoT data (see below).

Nick Millman, Accenture lead on big data and analytics, says mastering data management is the key component to getting the benefit of data coming from IoT devices.

“Organisations have got to have a clear strategy to drive value from IoT data – they need to be thinking about how to productionise this data,” he says. “To do that, you need MDM.”

If Gartner is right about the number of internet connected devices, business data will soon be dwarfed by IoT data. Organisations with the infrastructure ready to exploit the data will benefit from this revolution, but the rest could be left behind.

Case study: IoT cuts waste from rubbish collection

Rova is a Dutch public waste collector serving 21 communities across the Netherlands, where households dispose of waste in nearby containers. Before last summer, Rova used to empty 1,500 of these containers using vehicles travelling a fixed route.

Dennis de Jong, Rova's IT manager, says the average container was only about half full each time it was emptied. The company knew it could make do with fewer trips, but did not want to risk containers overflowing with rubbish.

Wheelie bins were already fitted with RFID chips – used to bill people for the rubbish discarded, encouraging recycling – and garbage trucks already had GPS. The collection containers themselves have access control devices allowing householders to open them with an ID card.

By putting containers online, Rova was able to collect data about how often they were used and download the information twice a day to specialist ERP software from GMT, using the OpenEdge platform developed by Progress Software.

Data analytics helps predict when each container is likely to be full. Route optimisation software guides the garbage trucks to the containers that most need emptying, avoiding those that are not yet full.

“Now each container is up to 75% full when we empty it,” says de Jong. “This means we have fewer vehicles, driving shorter distances, producing less CO2 by using data that is already there, but in a smarter way. We save about 20% of our operating costs using this software.”

Case study: Milton Keynes uses IoT for parking puzzle

Milton Keynes is the UK’s fastest-growing local authority. But with such growth comes problems. Traffic is set to increase by more than 5% in the next few years, only half of which can be accommodated by the town's infrastructure.

But more intelligent management of traffic and parking might help Milton Keynes cope with this volume of traffic without laying new roads and reducing the number of new parking spaces needed, according to Alan Ward, head of corporate ICT research at BT.

BT is a partner in MK:Smart, a £16m collaborative initiative partly funded by the Higher Education Funding Council for England and led by the Open University. It aims to employ technology such as IoT to aid economic growth in Milton Keynes.

BT’s initial project addressed the problem of parking. Using narrow-band radio transmission from nWave to lower cost and power consumption compared with 3G and 4G networks, BT is linking a variety of sensors to detect whether a parking space is in use.

Data will be uploaded to a BT data hub and become available to data scientists and application developers. The aim is to predict which parking spaces will be vacant and guide drivers to them. At the moment, the scheme is testing a sensor technology based on infrared and modulation of Wi-Fi signal when a car is in the space.

Data from a small number of parking sensors is live and online, but the project is looking for partners to commercialise the service.

What IoT means for data management

Use standards where available

Alan Ward, head of corporate ICT research at BT, says some standards will be useful, such as Standard ML, Extended ML, REST (a simple stateless architecture that generally runs over HTTP), JSon (JS Object Notation, a text-based, human-readable data interchange format). Other embedded devices will have sector-specific semantics that may need some bespoke code to bring into general-purpose computing.

Suppliers will help, but need partners

Paul Hingley, Siemens' data service business manager, UK and Ireland, says established data management suppliers such as IBM, Oracle and Teradata want to play a role in IoT data, but will need partnerships with suppliers that have a history in industry specific embedded systems.

Get ready for an explosion in volume

Nick Millman, Accenture's lead on big data and analytics, says: “There are big implication for the data environment for organisations to exploit the IoT. We are talking about much larger datasets than BI or operational reporting. Once a device is connected, it can produce continuous data in real time. You need the right platform, tools and processes to aggregate that. Then you can get value from data science.”

Look to extend data governance to IoT data

Gartner research vice-president Ted Friedman says: “It’s a big worry see suppliers with a specific IoT platform or solution. You worry it will be siloed, a thing off to the side. Our guidance is that although this stuff is new, you need to think about extending practice on data governance to IoT. Likewise, with analytics, you have made good investments, so figure out ways to leverage them.”

Read more on IT for manufacturing