AI, the key to useful data exploitation

The Internet of Things (IoT) and connected objects themselves are playing an increasingly important role in the everyday lives of users. For Patrice Slupowski, Orange’s Director of Digital Innovation, the data generated by this multitude of objects cannot be processed without artificial intelligence (AI).

“What are the 4Vs of Big Data? Volume, velocity, veracity and variety”

The IoT sector has seen significant growth, partly due to the extent of business investments (expected to reach 62 billion dollars in 2018, according to a study carried out by IDC at the end of 2017). This is accompanied by the inclusion of a growing number of connected objects in our daily lives, which generate a huge amount of data. This data represents a veritable gold mine for developing new services – as long as it’s processed and exploited in line with ethical, privacy and security regulations. For Patrice Slupowski, Orange’s Director of Digital Innovation, this formidable growth lever can only be pulled by using another form of technology: artificial intelligence.

How does the Internet of Things relate to artificial intelligence?

Patrice Slupowski: The Internet of Things is based on the connection of a plethora of objects that operate as a network. It will lead to the production of enormous quantities of raw data. This data must be qualified and processed into usable information. To achieve this, the use of AI seems crucial. Tools such as algorithms, or even learning mechanisms, like deep learning, will allow us to process, analyse and ultimately make the best use of this bulk of data.

Can you give us a practical example?

Patrice Slupowski: We have worked for the Sanoia and AP-HP (Paris public hospitals) e-health platform. We have contributed to the analysis of activity data for patients with rheumatoid arthritis. This is a very debilitating disease that affects 1% of the population and causes flare-ups. We’ve given actigraphy bracelets to 150 patients. They work like connected wristbands sold commercially for well-being or exercise purposes. Through the development of specific AI and the exploitation of data, we were able to identify and predict the onset of these flare-ups. Eventually, the aim is to change how the disease is treated, making it much more accurate.

A combination of data and AI will help create smarter IoT solutions. In your view, will data become the gold mine that we’re all expecting?

Patrice Slupowski: The promises around data are related to the ability to re-engineer all human activities by building data-driven models. The speed at which data is produced, and the increased power of micro-processors, suggest that we will see great strides in many areas: health, energy, business, etc. We will be able to understand the links between different variables and design apps with previously unimagined capabilities.

What type of data are we talking about? Can it all be collected and exploited, from both a technical and regulatory perspective?

Patrice Slupowski: It is important to distinguish between anonymous data and personal data. With regard to the former, we often talk about the 4Vs rule in the context of Big Data: volume, velocity, veracity and variety. These are fundamental to the work of data scientists. Of these four priorities, veracity is probably the most important point. This means ensuring the quality of the data to be exploited.

And with regard to personal data?

Patrice Slupowski: The most important concern around this data is privacy. From 25 May 2018, the GDPR (General Data Protection Regulation), which resulted from the work of the CNIL (French Data Protection Authority) and its European counterparts, will establish a harmonised framework at a European Union level, and grant new rights to citizens with regard to the collection and use of personal data. This document highlights a founding principle: personal data fundamentally belongs to the individual. This means that, from 2018, it’s the individual who must be in control of his or her data.

Returning to the issue of AI and IoT, what are the advantages of such an alliance?

Patrice Slupowski: AI’s ability to beat humans at chess or a game of Go is an impressive example of its aptitude, but that’s not the point. It’s very likely that data will become an asset in a huge number of areas with far greater relevance from a user’s point of view. AI and the IoT will help users to become better acquainted, to simplify their digital life, to discover new experiences, to protect their privacy, etc. Businesses won’t be left behind: this intersection will lead to gains in productivity, a better understanding of their customers, a transformation of their operations, new business ventures, process improvements, etc.

Sometimes, this new digital potential causes people to worry about the role of humans alongside the devices of tomorrow. What’s your view on this?

Patrice Slupowski: The anticipated progress around AI is generating many expectations and promises, but also a few delusions. We’re talking about radical transformation scenarios, which have captured the imagination of those who enjoy reading science fiction for many years.

In actual fact, we’re currently dealing with so-called ‘weak’ AI. It’s able to problem-solve in four main areas: understanding language, advanced data analysis to identify links, real-time data classification (for example for computer vision) and winning games with defined rules, like chess.

Humans are still relevant in many ways, and will be for a long time! This is particularly true in terms of our ability to express and recognise emotions. Advocating the role of humans in digital development is central to our efforts at Orange, whether in the day-to-day running of our company, in policies for developing our employees’ skills, or even in listening to our customers.

Read also on Hello Future

LightOn, a private and secure AI for businesses – Laurent Daudet


Sensors and AI ensure the good health of bridges


Using AI to monitor movement promises to improve home care


AI breaks the sound barrier


Fake news arms race prompts boom in AI


Big Data, algorithms and AI, the tools of tomorrow’s insurance?