The role of data analytics in the Internet of Things
Who would have thought that the Internet of Things (IoT) could impact anything from artificial insemination of cows to supply chain management decisions?
Everyone in IT is talking about the IoT. It has the potential to help organisations react quickly – even in real time – to events or factors that might impact businesses.
Take the example of the ‘Connected Cow’, a project where Microsoft, in partnership with Fujitsu, used activity trackers on cows to determine when female cows go into estrus and are ready to mate.
The data generated by the ‘fitbits for cows’ was analysed to determine when a cow was ready for artificial insemination, with an impressive 95% accuracy rate.
The project improved calf production by up to 31%, with an average of 12% across selected farms. You can watch how the project evolved here – www.youtube.com/watch?v=oY0mxwySaSo
The applications of the IoT are a lot broader. IoT refers to a network of ‘things’ (objects, sensors, devices, vehicles, components, people, animals, potentially anything) that are able to identify themselves and report on factors in their environment at a very granular level without human involvement.
In order to make an IoT estate work for an organisation, it is essential to combine Operational Technology expertise with Information Technology expertise.
An example implementation given by Microsoft using their Azure cloud tools involves a fictitious transport company moving temperature-controlled cargo. Using sensors on a vehicle, the company can monitor the state of the cargo and the vehicle itself.
This data can be integrated directly with vehicle communications and tracking systems in order to handle exceptional cases such as refrigeration failure.
However, while data collected from an IoT estate can be invaluable in transforming the shape of an organisation, due to its highly granular nature it also has the potential to drown an organisation in so much data that the value is lost.
The data collected from IoT devices is often in differing formats and of such a volume that it can catapult a company into the realm of Big Data. It can be a daunting and difficult task to process all this data and blend in reference data from within an organisation in order to drive useful insights.
Furthermore, insights gained from the analysis of IoT data need to be surfaced to business decision makers and operational managers so real business benefit can be realised and decisions made, whether these be strategic or operational in nature.
Cloud solutions are tackling these issues offering almost unlimited scalability for collecting data. Running queries in an ever-simplified manner across both structured and unstructured data means it is now quicker than ever to make sense of multiple data sources and large amounts of data from IoT estates and internal data sources.
These cloud solutions are making it easier to perform analytics on live data streams to provide immediate visibility of operational factors, as well as integrating machine learning to assist decision-making with predictive analytics.
Data Visualisation is essential to surfacing insights to end users and the tools available are numerous, providing us with the ability to explore data from an underlying data warehouse without having the technical knowledge of the data or programming involved.
Example of filtered analysis using a data visualisation tool to highlight shipment trends
Products available range from desktop to cloud to enterprise scale, and provide extremely rich sets of features to analysts for creating data-driven experiences for the end-user, and sharing insights easily in a consistent and powerful way.
With the right combination of technical Business Intelligence and analyst capabilities, organisations can ensure end-users benefit from up-to-date information collected at a very granular level to drive efficiencies.
The author is Nick Finch, technical director at data analytics firm, Concentra
Nick Finch is responsible for technology across product development and client delivery, focusing in particular on the junction between Analytics, Technology and Cyber Security.