SAS has had extensive analytical software for many years, but to what extent is the software changing as a result of emerging influential technology such as artificial intelligence (AI)? AI makes it easier to draw conclusions during data analysis. We discussed AI’s influence on analytics with Mark Bakker, Principal Business Solution Manager, and Matthieu Joosten, Industry Leader Telecom South West Europe.

According to Joosten, there is some confusion about AI in general, with which he more or less refers to the extensive possibilities of the SAS Viya platform. When companies talk about AI, they regularly apply one or more AI-related issues. This means that they sometimes see Business Intelligence (BI) or Robotic Process Automation (RPA) as their major AI application.

SAS Viya clearly offers more and aims to cover the entire analytics lifecycle. The platform is an enabler of different AI disciplines, such as image recognition, machine learning, neural networks and algorithms. Users use these methods to analyze and calculate certain situations.

Practical examples must convince

During our conversations, the unique situations in which AI can prove its worth are emphasized. Bakker really sees the added value in this, since the usefulness of AI is demonstrated in specific business situations. Seeing is believing, or seeing is believing, as it is also described. If, in a similar situation, it is demonstrated that artificial intelligence offers added value for employees or that it is socially relevant, then, as a company, you are inclined to follow the example. But enthusiasm is also stimulated if the added value quickly becomes clear.

According to Bakker, the challenge does not necessarily lie in the development of technology. After all, AI has been in development for some time now, which means that it already has quite a few possibilities. Those possibilities just have to be well demonstrated. As a result, there are already situations in which SAS is making a pre-investment, in order to deal with further commercialization at a later stage. SAS does, of course, insist on the latter, but the drive of SAS to give AI a big push is great.

Preventing problems by means of AI

One of the cases in which the company demonstrates the possibilities of AI is aimed at telecom operators. SAS uses its software for these parties to predict, on the basis of analytics, when a certain network element will fail o that the end customer does not have a problem. Artificial intelligence is then the overarching term for detecting the problems, which is often possible because a pattern indicates a certain error that will occur. This occurs in a modem, for example.

Different technologies are used to detect problems. An important component in this is streaming analytics, also known as event stream processing. An event is then the event that takes place at a certain moment in time; the data stream is the continuous flow of data events and the processing section analyses all data. In addition, machine learning models are also part of the approach, which automatically determines whether the problem solves itself automatically or whether it requires action.

Image recognition as an AI application

Another test of SAS focuses on image recognition, a technology in which the company invests heavily, in combination with 5G. Together with an Italian telecom operator, the network standard is used to analyse images from a square. 5G is important in this situation because the bandwidth is larger, and the latency is less, which makes real-time analytics possible.

In this specific trial, cameras were placed on a square to monitor the crowd density (number of people) and the movement of the crowd. The SAS software is therefore not used for an extensive analysis of the individual, which is sometimes the subject of privacy concerns. It is purely meant for crowd management and security so that emergency services are supported in the event of panic and attack situations. Image recognition then determines how many people are on the square and what the best escape route is.

The trial also uses various AI applications. Real-time analytics assesses how many people there are and how threatening the situation is. On the other hand, SAS uses a piece of machine learning to learn from the number of people. From the movement of the number of people, certain conclusions can be drawn. Another example of an aspect that monitors machine learning is the behaviour of a teenager or an adult in a group, which means that conclusions can also be drawn from this for better crowd management.

The importance of AI is becoming increasingly clear

With the user applications, SAS is working on mapping out the AI possibilities. This dedication was recently demonstrated by the announcement of an investment of 1 billion dollars (880 million euros) in AI. It should benefit Research & Development (R&D), focusing on simplifying the benefits for users with different skill levels. In addition to data scientists and data analysts, it also applies to business users, for example. The SAS software also needs to be upgraded through the investment, something that will become clearer in the coming years. The investment will be spread out over three years.

In this way, SAS is working on improving and making artificial intelligence more widely available. This is not surprising since most analytics applications are already very focused on AI. We are curious to see how these AI functionalities will continue to develop.