Tackling the connected future, the right way

The Internet of Things (IoT) is the latest IT topic to receive a consistent wave of industry attention, much like Cloud Computing and Big Data did before it. There is considerable discussion about the potential applications of IoT, with excitement about what the next connected device on the market might be and the types of new applications and revenue models that might follow.

Estimates about the growth opportunities in the connected devices market vary – with Bosch SI estimating that six billion devices will be connected by 2015, Gartner predicting 26 billion connected devices by 2020, and Cisco estimating 50 billion devices will be connected by 2020.

A survey by The Economist Intelligence Unit found that 75 per cent of companies are either using or exploring the IoT, with IDC estimating over $7 trillion (£4.1 trillion) in economic value added by the IoT What's clear – the technology is developing quickly and is well on its way to being a staple in our personal and professional lives.

This is obviously a significant opportunity, and the number of telecoms operators offering dedicated network solutions for Machine-2-Machine as the foundation for IoT is testament to this. But as the number of connected devices increases, and the use cases diversify, businesses need to ensure they have the structure in place to deliver long term benefits.

The Internet of Things represents a considerable opportunity, but businesses need an infrastructure that affords the flexibility, agility and scalability to meet the changes that this new era will bring. The Internet of Things will evolve and the infrastructure around it will need to keep pace.

Retail leading the way

The retail industry often leads the way in the adoption of new technology. Innovation is key to the delivery of successful customer service and in such a competitive market, technology is often a clear differentiator. One piece of technology, Radio Frequency Identification (RFID) has been used successfully for years to track containers, pallets and crates, primarily in closed loop systems and mostly with high-value goods.

The increased sophistication of IoT technologies means that costs are reducing for RFID and similar tools, making the tracking of goods on an item-level a feasible business case. For retailers, this will have many advantages, including inventory accuracy, reduction of administrative overheads, automated customer check-out processes and a reliable anti-theft system.

However, to successfully track individual products, software investments need to be made to manage the data being produced by all connected components. Traditional databases created in the 1970s were not developed to handle the complexity, volume or speed of data that comes in structured, unstructured and semi-structured forms, with the ability to blend it across multiple business processes and analyse it in real time to drive automated operational processes.

This is something that only becomes more apparent as retailers begin innovating with IoT. For example, in the future someone entering a supermarket might have that particular store's app up and running, integrating with the store through an indoor positioning system such as a beacon. Their shopping list will engage with the store's app, a map will be displayed and the app will notify them when the customer is near a product they need.

Such a scenario is only a snapshot of how the retail industry will evolve with the integration of IoT, and illustrates the complexity of the processes involved. Such intricacy requires a database that can cope with the growing demands placed upon it, as well as any changes or innovations the user wants to make. This is without even considering how the data produced by these interactions will be analysed.

Big data integration

As IoT develops, where the real value will be delivered to businesses is through the data itself. The vast amount of rapidly changing information created by IoT must be analysed in real time to allow applications to deliver enhanced operational insight.

This need to accommodate an ever-increasing army of sensors and to analyse this data in place presents new demands on all components of the technology stack –particularly in the underlying databases used to store, manage, process and analyse the data.

Without the appropriate technology stack in place, IoT will simply produce data within siloes and won't receive the analysis to drive insight. The traditional databases in place were never designed to offer the needed flexibility or scale in the same way as non-relational databases, whose dynamic and distributed architecture allows them to adapt over time.

Delivering long term

We are only at the very beginning of the interconnected future and over the coming decades IoT will become integrated into every dimension of our lives.

This is a tremendous opportunity and will transform the way we live and do business. The companies that will succeed and reap the most benefit are those that appreciate that IoT will evolve over time, and account for this in their infrastructure.

Those organisations that bury their heads in the sand, ignore the coming reality and continue to invest in existing and outdated infrastructure will lose out. It's always difficult to predict what's coming around the corner, but the best strategy is to ensure that there is built in flexibility to fully exploit the coming wave of interconnected devices and assets.

Mat Keep is principal product marketing manager at MongoDB