Of all the many and varied changes in technology that have emerged over the past decade, the Internet of Things (IoT) is potentially one of the most impactful. In fact, CompTIA’s Emerging Technology Community has stated that IoT is the most impactful, while statista has predicted that by the end of 2019, there will be 26.66 billion IoT devices, rising to 75.44 billion by 2025.
Not only are these devices all connected to the internet, but many of them will also be talking to each other. The sheer volume of data that will come from all these billions of devices is one factor that needs addressing for organisations hoping to make sense of it, but also the actual relevance of that data. When you consider how hard many organisations find the mere storing of content and documents, figuring out how to do that for IoT data is going to be a major challenge.
Given this growing importance and volume of IoT data, what should organisations be mindful of when managing it and what’s the best approach to adopt when attempting to address this challenge?
The rise of IoT…and a range of associated challenges
It feels like we are not too far away from a situation in which almost every single device, appliance or gadget can be connected to the internet. Gartner has previously predicted that by 2020, IoT technology will be in 95 per cent of electronics for new product designs and that’s a figure that is likely only going to edge closer to 100 per cent over the next few years.
For consumers this offers a number of benefits - ever more personalised products, a deeper insight into health and fitness, greater convenience and ultimately a much better user experience. For businesses, it means yet more data - much more data. Given that we live and work in an era where the volume of data generated is growing faster than ever before, that is quite a prospect. In 2018, IDC's ‘Data Age 2025’ whitepaper predicted that the collective sum of the world’s data will grow from 33 zettabytes in 2018 to a colossal 175ZB by 2025.
Managing all of this IoT data is set to be a major task for many businesses and the rise and growth of IoT data comes with a number of associated challenges. Many organisations will lack the architectures, policies and technologies that address the full data lifecycle. Current approaches and infrastructures will need to be overhauled and / or scaled in order to get the most from IoT data.
There is also the question of immediacy with IoT. Data is generated so quickly and has such a short shelf-life, storage becomes a problem. IoT is dependent on fast data and immediate insight, and connecting a wide range of devices can make real-time processing and analysis that much harder. The IDC whitepaper stated that by 2025, 30 per cent of the data generated will be consumed in real-time. How will all this data be recorded for compliance and legal purposes? What data must be recorded before and after the fact to prove that the right decision was made?
Finally, we are now more than a year into GDPR and organisations need to demonstrate that they take appropriate care with every single piece of data coming into the business. IoT adds a significant layer of volume and complexity to this process. The penalties are severe for any organisation that is non-compliant with GDPR – already in 2019 we have seen a number of record fines for non-compliance - and it’s an issue that must be factored into discussions about data management and IoT for any business keen to avoid such a penalty.
The importance of metadata
What cannot be overlooked regarding IoT data, is the value and importance of metadata. Any data set is only truly valuable to an organisation when viewed with context and relevance – this applies to IoT data just as much as it does with traditional enterprise data and content.
This means that metadata – the data that provides information about other data sets – become even more important. One example would be a sensor that is continuously uploading data. This means nothing unless you know what that sensor is recording, what the default values are, what the extremities are and how it relates to other sensors. Data needs to have context to provide business value.
Any organisation that wants to maximise the value of its IoT data must label, categorise and describe that data effectively, and this is where a modern enterprise content management (ECM) solution provides value. Next-generation ECM systems now employ cutting-edge artificial intelligence (AI) and machine learning technologies to identify data and content and then automatically apply the appropriate set of metadata values.
But to be truly effective though, it needs more than a generic AI solution. It requires something more sophisticated, and context aware. Organisations must able to train their own custom AI models using business-specific IoT data sets. This delivers much better results, with the metadata management constantly improving as the machine learning develops.
In addition, AI also allows an organisation to address IoT interoperability. Each device that connects to an organisation must be able to communicate and exchange data with other devices that do so. This adds the greater layer of context that is so important in maximising the value of IoT data. An AI-enabled platform can easily identify any new device by looking at the model numbers and other classification data. Once this has taken place, that device can then be connected to any other devices and data can be exchanged. Metadata can even be added to the exchanged data, so devices can all process that data correctly too.
As the volume of it grows, IoT data affords a massive opportunity for insight and value but there is the possibility that many organisations will just be completely overwhelmed. The smart use of metadata and deploying the latest AI technology can help ensure they won’t be.
- The need for next-generation data management: Six key storage considerations for healthcare organisations
Uri Kogan, Vice President of Product Marketing, Nuxeo