Machina: The value of M2M is in the (big) data

By Emil Berthelsen and Matt Hatton

Matt Hatton, left, and Emil Berthelsen

There has been a lot of discussion recently about Big Data and the impact on machine-to-machine (M2M). This month Machina Research published its report "Creating value from data analytics in M2M – the Big Data opportunity" and in this article we pull out a few of the key themes of the research.

The core of what we have known as M2M in the past has focused on device-centric solutions requiring the transmission of basic data. As an example a simple home alarm device needs only report that it has been triggered. However, as the industry matures we increasingly see a shift from the simple transmission of data to two-way communication between platform and device to create intelligent feedback processes. It also involves the provision of richer information, based on pulling in numerous additional data sources. A good example of this would be usage-based insurance (UBI), where the application pulls in behaviour data from the driver and analyses it to determine the risk profile of their driving. This contextualisation of data requires a richer set of capabilities and shifting from simply carrying data to providing risk-scoring based on that data is also the key to unlocking a tremendous amount more value for the provider of the service.

In our latest report we identify a number of key factors that are accelerating the growth in Big Data analytics in M2M. Low cost devices, cheaper data plans, new network technologies and all the great work done by carriers and vendors in driving out the complexity from M2M is reducing the barriers to adoption. As a result more and more potential data sources are becoming available. On the IT side of things new data processing technology (e.g. Hadoop, NoSQL), cheaper data storage (including cloud-based solutions) and cheaper processing are all opening up the potential to crunch all of this newly available data.

Big Data is not simply about the volume of data that needs to be managed. It is the interplay of five factors: size, speed, structure, situation and significance (where 'significance' is derived from data through analytics). What distinguishes Big Data from traditional Business Intelligence (BI) analysis is the sheer scale and complexity of the data as well as the highlighted immediacy (speed) of using the data for further action and/or decision-making. Big Data analytics in M2M requires the management of huge amounts of structured and unstructured data (as well as metadata), managed in real-time, and processed to deliver meaningful and useful insights of significance and value. This significance can be measured in terms of a Data Significance Factor (DSF) which signifies the predicted importance of the data at the analytics processing stage. It is influenced by factors such as the requirement for prioritisation, the impact of the event, and predictability of the data. The DSF of a healthcare remote monitoring application, for instance, is higher than for a smart meter given the immediacy and seriousness of the potential impact of the associated data.

While it is a tremendous opportunity, Big Data also raises a number of social, business and technical questions and challenges. Enterprises, will need to break down artificial walls and divides, and there will need to be extensive testing and proving of the benefits of combining multiple data sources within the business. For consumers, privacy is the critical issue, and enterprises will need to work hard at providing transparency around the management of consumer data and ensuring that the correct level of privacy restrictions and requirements are addressed. To deal with all the relevant issues here would require an article in itself, but safe to say that all parties involved in Big Data analytics are (or should be) examining issues such as codes of conduct, privacy by consent, and fair usage.

The value of Big Data emerges from delivering integrated data analysis of both internal (enterprise) and external (public and social) data,  predominantly created from M2M solutions deployed within identified data communities but also from other information sources from the data community partners. The challenge in M2M and Big Data is establishing a framework whereby data takes priority as it becomes an enterprise asset with value.

In truth, all of the discussion around Big Data in M2M misses the point. In truth the two are inseparable. Without data management there is no value in M2M applications. The conclusion is that the value of M2M lies predominantly in the management and analysis of data and those companies hoping to generate serious revenue from M2M will need a Big Data strategy. It should also be noted that this mashing-up of various data sources is very closely connected to the development of the Internet of Things (IoT) which drives the combination of numerous sources of information. The road from M2M to the true IoT necessarily involves data analytics.

Machina Research provides strategic advice on the newly emerging M2M, IoT and Big Data markets. For more information on Machina Research and the report, click here. Matt Hatton is director at Machina Research, [email protected], and Emil Berthelsen is principal analyst at Machina Research [email protected].