The Fourth Industrial Revolution is underway.
Technologies such as robotics, nanotechnology, the Internet of Things, AI and machine learning are forming "cyber-physical systems" that will transform the world as we know it.
For many industries, some of these technologies still seem remote and futuristic. Do they really affect your business? Can you hold off a bit longer, or should you look at deploying them now?
The Internet of Things (IoT) is the concept of extending internet connectivity beyond computers and mobile devices, to traditionally non-connected, everyday objects. These range from cars and household appliances to health trackers and smart sensors embedded anywhere from buildings to roads. The IoT already has billions of connected devices, and will have over 200 billion by 2020, according to Intel. All of these devices are able to collect and transmit vast amounts of data.
Managing Big Data
This harvesting of information results in Big Data. Managing it is challenging - it consists of data sets that are too large and complex for traditional analysis methods or software. You can’t put Big Data into an Excel spreadsheet and have a human crunch the numbers. It has to be handled by computers running advanced data-processing software.
Using data analytics, these computers clean up and organise data, and analyse it to show patterns and anomalies that can provide insight into a process or situation. Netflix has over 140 million subscribers, and analyses viewing patterns to see who is viewing what, and when. It can then use this to fine-tune its Recommendations system and plan new content. PepsiCo collects warehouse and POS data from all its clients, and uses this to forecast production and shipment needs.
But there are limitations to data analytics. Data initially shows you the "what", not the "why". For example, the data might reveal a problem, such as a huge fall in customers in a specific location. However it can’t give you the reason - which you need to know if you want to fix the issue. Qualitative analysis - by a human - is still needed.
The role of AI and machine learning
The next step is using artificial intelligence (AI) and machine learning (ML) to autonomously run processes based on the data analysed. There are varying definitions on what exactly constitutes AI. Researchers Andreas Kaplan and Michael Haenlein define it as “a system’s ability to correctly interpret external data, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation."
Machine capabilities generally classified as AI include tasks such as understanding human speech, competing at strategic games like chess and autonomously operating cars, or being able to predict failures and trigger maintenance processes autonomously.
AI starts with human-created algorithms: instructions or rules on what to do depending on the data input. As a simple example, an email program might have rule-based filters that send certain emails to the Junk folder, e.g. specific keywords or subject headers.
Machine learning takes this a step further. Without additional human input, an app uses data to improve its algorithms and apply better rules. It might detect that a user frequently marks certain kinds of email as spam, and "learns" that these should in future be automatically sent to the Junk folder. Google uses machine learning to improve Gmail’s spam filters, looking for new patterns that might suggest an email is not to be trusted. This could be anything from the formatting of an email to the time of day it’s sent.
One example is Doxel, which uses AI-enhanced software to monitor construction sites, using robots, camera-equipped drones and LiDAR sensors. It processes visual data with deep learning algorithms, and can detect errors in construction by comparing data from everyday scans to small scale design models.
In Asset Management, machine learning may come in to play when looking at how areas with similar footfall degrade over periods of time, based on external influences as temperature and humidity. Over time, predictive modelling is improved based on previous ‘experience’ with various events.
Boeing has been combining IoT and blockchain to track the provenance and lifecycle of aircraft components. The IoT provided a quicker, deeper understanding of a machine, and the company can then apply predictive analytics to anticipate when there’s going to a problem. This helps fix issues in advance, as well as extending the life of components. It also helps with forecasting demand and rate of consumption.
Industry 4.0 is significant for every business and every industry. Big data analytics is an important investment for all organisations. As well as your own data, there are public data sets out there, such as Google Trends, various government repositories of social, economic and population data, and social media monitoring tools. Having a strategy and a roadmap for Industry 4.0 will be critical to survive and compete in this revolution.