Working with ‘Big Data’: Insights and benefits with Pradipto Biswas, Data Science and Data Insights Lead
Pradipto Biswas joined Sagentia in January 2021, continuing an illustrious career spent working primarily in global financial services. Pradipto initially trained as an engineer, but his interest in building information and smart design systems quickly led him to specialise in data and analytics. This led to roles with internationally renowned finance and professional services firms – including PwC, Lloyds, Accenture and Deloitte – as well as further academic studies and a fellowship at London Business School.
Pradipto is now using his expertise in Data Science to apply what he’s learned in finance and mathematics to more tangible applications for Sagentia, in industries like oil and gas, as well as medical devices, food science, agriculture and elsewhere. Here Pradipto, Data Science and Data Insights Lead, explores the benefits Data Science can bring to a project, as well as some of the trends that are shaping the growing appetite for, and understanding of, ‘Big Data’.
Why is Data Science important?
Data is a bit like the new oil of the economy, it’s powering transformational changes which can make businesses more efficient, effective, and competitive in the marketplace. Data Science draws on techniques from mathematics and statistics, combines them with data and analytics, and brings all this knowledge together to solve concrete challenges, rather than purely academic problems. That’s what I find so exciting about it.
Data Science is an area of technology which has the potential to radically transform the way our clients operate. This transformation has already started, but what have seen so far is just the beginning. These changes are radical, rather than incremental, in nature and are often collectively referred to as the ‘Fourth Industrial Revolution’. While that may seem like a big claim to make, businesses will need to adapt and adopt Data Science techniques and technologies if they are to survive and thrive in this new world.
How can we use Data Science?
We can use Data Science to find insights, patterns, and correlations that the human mind alone would be unable to see. People are not great at handling and processing large data sets, but this is an area where a computer’s capabilities can complement our own. Data Science techniques, combined with data visualisation, can cut through the noise and highlight interesting patterns to deliver useful insights.
One practical application might be analysing huge volumes of GPS data to track vehicle movements. A computer can analyse this data to deduce information such as speed, direction, location and proximity, visualising interesting and unusual patterns. These techniques can be used to analyse a ship’s transponder data, including periods when it ‘goes dark’ by switching its transponders off. We can then use the ship’s last known location, speed and bearing to extrapolate where a ship may have gone during this period – information that’s useful in detecting enemy military manoeuvres or smuggling activity, for example.
What else can Data Science help us to do?
Behavioural modelling is a particularly useful application of Data Science for studying and understanding patterns of behaviour. We can use it to ask questions like ‘what is normal versus abnormal?’, in a given context, and ‘how abnormal is it?’. These answers may provide all kinds of useful information. We can look at people’s behaviour and group them in segments to come up with highly targeted products, services and ways of interacting with them, based on their individual or group preferences. We can also use this information to automatically flag abnormal events, based on unusual patterns of behaviour, which could have potentially disastrous consequences – such as a patient in need of urgent care or the start of a new type of cyber-attack.
Predictive modelling is another interesting area with similar potential for avoiding unwanted outcomes. We can use these techniques to examine past data and build models to predict future events and intervene appropriately. In practice, this might be by predicting when a costly piece of machinery is about to fail, allowing us to carry out predictive maintenance and save time and money wasted on outages.
How can we cut through the hype?
Data Science – much like Artificial Intelligence (AI) and Machine Learning – is a bit of a buzzword at the moment, and there are consequences to all of this hype. A key risk is that these technologies, despite their tremendous potential, remain stuck on the wrong side of the innovation curve if we fail to manage them properly. We need to have realistic expectations about what these technologies and techniques can and can’t achieve, and there needs to be wider realisation that they are not magic wands. There is no alternative to sound architecture and planning – based on an in-depth understanding of these technologies, their capabilities, limitations, and drawbacks.
The hype has resulted in a lot of investment – both from product vendors looking to ‘AI-enable’ their platforms, and companies looking to use AI to improve their performance – yet most initiatives remain limited to experimentation and constrained to the lab. Most companies are not taking the next step, to operationalise models developed in the lab. This is largely uncharted territory, requiring radical changes in business processes and the way people work to properly achieve benefits.
What does Data Science hold for the future?
Some of the techniques popularised by Data Science are likely to become even more widely adopted in the near future, particularly as the Internet of Things (IoT) becomes bigger. The volume and variety of telemetry data provided by the increasing numbers of autonomous and non-autonomous devices is too much for traditional analytics to handle, so there’s going to be a big push in Data Science driven by manufacturers and users of IoT devices.
The bulk of investments and developments in Data Science over the last few years have been driven by banks and other financial services institutions. Although they will continue their investments in these areas, I believe that the next revolution will come through industrial manufacturers, mining firms, agritech, and consumer applications, widening Data Science to incorporate a bigger range of data and device-based (aka ‘Edge’) computing.
Among the changes we’re likely to see is the increasing use of sensors, which will become more widely adopted as they continue to become more affordable. Sensors can collect and monitor a variety of information, and can be used for tackling things like predicting food spoilage and cutting down on food waste, or for patient care in medical settings, such as predicting when a patient is going to go into a critical stage and ensuring the necessary care is available. We can also use sensors to anticipate crisis points and intervene, rather than simply reacting.
Remote sensing can also make mining, agricultural, or oil and gas operations, smarter and safer. Data provided by these sensors could be used to help monitor oil rigs remotely, to ensure drilling and mining operations are situated in the best possible location, or to optimise farming methods. We’re already seeing data used in agriculture to power developments in smarter planning of how fertiliser is used, or soil data being used to plant crops more effectively and autonomously. Data Science will also be used increasingly to help detect and prevent fraud, cyber-attacks, financial crime, terrorist attacks, epidemics, and more.
How can Sagentia help its clients navigate this journey?
Significant changes are on the horizon that will undoubtedly lead to increased marketplace competition, in addition to growing demands on businesses from customers and governments alike. Businesses will need to adapt or fall behind and Data Science is already becoming a critical tool for navigating these developments.
Data Science alone is not enough, however, and businesses will need to partner these developing techniques and technologies with deep domain expertise, innovation and pragmatism. Sagentia brings together a history of innovation, powered by science and engineering, with long-established skills in medical, industrial, consumer, and food and beverage industries to accompany its growing Data Science and Data Insight practice.
We can help you navigate your end-to-end journey with a wide range of data-driven skills – including building a strategic roadmap, designing new products, building sensors and collecting data, identifying and acquiring data from the right sources, data cleansing, combining and enriching data, visualising hidden patterns, building predictive models, behavioural modelling, and uncovering unusual high risk events. Most importantly of all, we will work in partnership with you to bring these capabilities to bear on your latest project to ensure its success.