By Richard Slater, Principal Consultant, Amido
Predictions for how commoditisation through cloud computing will shape cloud services across all industries
The commoditisation of IT started with the cloud. With commoditisation comes standardisation, which allows for inexpensive products to work well together in the cloud ecosystem and that can be delivered at scale. It also opens the door to technologies like blockchain to enable distributed trust. These are the significant cloud growth areas we anticipate over the next 18 months:
Serverless/FaaS1 is a hot topic in the world of software architecture, but we have to remember that cloud computing enabled the creation of Infrastructure as a Service (IaaS). IaaS enabled the creation of Platform as a Service (PaaS), and PaaS, in turn, enabled the creation of Software as a Service (SaaS) and the emerging Functions as a Service (FaaS) or Back end as a Service (BaaS). This is most commonly referred to as Serverless computing. With the launch of AWS Lambda in 2014, the term started to attract more attention from outside the developer community. FaaS is a commoditised function of cloud computing, and one that takes away wasted compute associated with idle server storage and infrastructure.
The significance of FaaS for businesses could be huge, as they will no longer have to pay for the redundant use of servers. Instead, just for how much computing power that application consumes per millisecond, much like the per-second billing approach containers are moving towards. Instead of having an application on a server, the business can run it directly from the cloud allowing it to choose when to use and pay for it, per task – thus making it event-driven. According to Gartner, by 2020, event-sourced, real-time situational awareness will be a required characteristic for 80% of digital business solutions, and 80% of new business ecosystems will require support for event processing.
However, not every business is going to be right for FaaS or serverless despite there being a real appetite in the industry to reduce the cost of adopting the cloud. Businesses considering FaaS as an option must realise that this is the ultimate in vendor lock-in, as it’s not easy to move services from one cloud to another. Each cloud provider approaches FaaS in a different way and at present functions cannot be moved between vendors.
As the demand for serverless technologies grows, the nature of DevOps will change; we may be moving into a world of NoOps, where applications run themselves in the cloud with no infrastructure and little human involvement. Indeed, humans will need to be there to help automate those services, but they won’t be required to do as much coding or testing as they do now. With the advent of AI, the IoT, and other technologies, business events can be detected more quickly and analysed in greater detail so enterprises should embrace ‘event thinking’ and Lambda Architectures as part of a digital landscape.
IoT on the Edge
IoT has contributed to a serious rise in the types and amount of datasets generated and the proliferation of things connected to the Internet has meant that Edge Computing is hugely important for all industries. Research firm IHS predicts that IoT will grow to reach a staggering 75 billion devices by 2025.
Right now, there needs to be a way to aggregate, analyse and distribute that data from the ‘things’ and send it back to the ‘things,’ quicker. Currently, Edge computing technology such as AWS Greengrass collects data and processes it from nearby items, sending it back to the cloud where analytics and machine learning (ML) can take place in order to make sense of the data, before sending the data back to the edge and subsequently the things – making them more intelligent.
The next wave will be for the compute to move from the cloud towards the edge giving the objects the ability to make intelligent real-time decisions i.e. a car needing to make a split-second decision on whether it should apply the brakes to avoid an accident. Edge computing holds data analysed from the cloud that is immediately passed to the object for instantaneous updates and responses, and Cloud to the Edge data distribution is essential to improve customer experience in industries like manufacturing, health and retail.
Industrial Internet of Things (IIoT)
Today, there are roughly 6.4 billion data-communicating objects in the world and this number is forecast to triple by 2020 (Accenture). The majority of these objects will be ‘things’ – whether cars, white goods or industrial assets: aka smart machines. These smart machines – featuring a multitude of sensors, automation proficiency and machine-to-machine communication capability constitute the IIoT. IIoT will enable data-driven manufacturing, where process and floor-wide monitoring are able to optimise efficiency and quality, through the application of machine learning to big data. This is being heralded as the revolution that will introduce huge productivity boosts to the industry.
As the UK launches its first state of the art, fourth industrial revolution (Industry 4.0) factory – AMRC Sheffield Factory 2050 – the potential for Edge Computing within Industrial IoT is accelerating. Dedicated to conducting collaborative research into reconfigurable robotic, digitally assisted assembly and machining technologies, there will be a need for a high variation and mass customisation of manufacturing throughout a diverse range of engineering sectors. This will shorten lead times and optimise costs accumulated throughout the supply chain, and that can rapidly ramp production up or down to meet demand.
Big data technology processes large volumes of information, collected by sensors on each machine, cell and the building itself to enable automation – without forsaking the need for humans. Look at how Fujitsu, with its UBIQUITOUSWARE, takes in an immersed reality with its products that enable humans to do a better job with the use of real-time analytics and data collected from other scenarios. If you fell in the factory, or if there’s a potential danger, edge technology and sensors can feed and receive this information directly to and from the worker. Machines can cause harm – robots that know when humans are close and slow down to protect them are an important part of the Industry 4.0 or IIOT revolution.
With the evolution of IoT and the vastly larger data sets that are streaming into the cloud, it would be impractical to try and process that quantity of data in real-time. However, you can use AI and event sourcing to summarise and generate actionable insights. Gartner wrote that 59% of organisations are still gathering information to build their AI strategies, while the remainder has already made progress in piloting or adopting AI solutions.
For businesses wanting to make use of structured and unstructured data to stimulate intelligent decisions and spot trends across all departments, it’s time to focus on data engineering, data lakes and ML in a practical way to identify data sets that would provide the most benefit from building a machine learning capability. This could include things such as fraud or purchase recommendations and up/cross-sell. Furthermore, businesses need to adopt viable cloud services that will benefit the business in the long-term – and already conversations with organisations are shifting to discuss how all these technologies will enable better cloud performance without breaking the bank.