Sigma planning to outdo sherlock at logical reasoning
Everyone’s talking about AI and how it is the absolute latest in IT. It is true to a certain extent but it is rather its accessibility that has opened our eyes and given rise to a massive wave of interest in artificial intelligence, or automated machine learning as it was called it in the 60s and 70s, when it was being used seriously for the first time.Post |
Mattias Jönsson of Sigma IT Consulting welcomes us.
To learn more about AI and to hear how Sigma is currently using it and what plans they have for the future, we visit the Sigma Building in the old Kockums area in Malmö. We take the lift to the fifth floor and Mattias Jönsson of Sigma IT Consulting welcomes us at the door. We start by asking Mattias what is meant by AI and what the term covers. “Some would probably say that it has magical properties and can do everything but it is, in fact, just applied maths. It uses statistics, optimisation and simulation to calculate probabilities and draw conclusions,” he explains. It is evident that here we have a mathematician who is passionate about this, but at the same time he seems pretty calm about the prevailing AI hype.
So why has interest in AI exploded recently? Everyone is talking about it and what it can do. Mattias explains that AI is not really anything new. What is different now compared to a few years ago is partly that these days we are generating more data that we actually have a use for. We leave traces all the time today. When we pay for parking with an app, data is generated. When we buy food, data is generated. The amount of useful data has increased and being able to gather it in the right place in real time, by using cloud solutions, also plays a considerable role in the growth of AI. “We collect an amazing amount of data today. It gives us and our colleagues in the industry so much more raw material to work with,” says Mattias.
"Some would probably say that it has magical properties and can do everything but it is, in fact, just applied maths. It uses statistics, optimisation and simulation to calculate probabilities and draw conclusions."
One decisive factor is of course the availability of computer power. Previously there were just a few “supercomputers” in existence that could make these advanced calculations. Nowadays, if you don’t have access to a “supercomputer” you can gain temporary rental access to computer power which will enable you to connect to the nodes that are needed for the duration of your calculation. This, in combination with intelligence and access to more data, has revolutionised AI and now more people than ever are able to optimise production planning, various business scenarios and make their logistics flow in a whole new way. Today, with the help of AI, we can optimise and streamline the competence that has always existed within companies. It makes processes millions of times faster and more accurate. Peter in the warehouse no doubt knew precisely at which times of the year there was more space available and which products left the warehouse faster than others. With AI, those facts and his knowledge become a formula that the company can apply to streamline its stock, minimise costs and maximise revenue, and be able to do it in a fraction of the time. So, as you see, AI is not only about making robots and teaching them to teach themselves. It can be about ordinary everyday things like logistics and warehousing.
Mattias emphasises the importance of not being carried along by the AI hype. Instead he recommends tackling it with a thorough before and after analysis because, without a strategy, there is a risk that an AI-initiative will be more expensive than necessary without giving you the gains that you were hoping for. “It is not AI you need, it is the solutions you come up with when using it,” says Mattias enthusiastically.
When we ask Mattias whether Sigma has ready-made AI solutions that they can offer their customers, it turns out they do. An example of this would be solutions that learn to detect deviations in machines or other production units. “We have some ready-made solutions, but since the role of analysis and strategy are so important for the outcome, we almost always create a tailor-made solution for each individual project or company,” continues Mattias. Even though some customers come up with a carefully thought through plan for how they want to use AI, they still bring in Sigma at an early stage. This means that Mattias and his colleagues can examine the customer’s needs right at the start and thereby determine whether AI should be part of the solution.
"It is not Ai you need, it is the solutions you come up with when using it."
“Some people maybe think that AI seems complicated and costly, but in many cases it’s just about simple, run-of-the mill things; simple but clever solutions which facilitate, improve and increase profitability,” explains Mattias when we ask him if AI is all about super complicated computer calculations. AI can be used to collect data so that things you previously had ideas about and opinions on can now be processed into facts. Most common is some form of automation, but AI can also assist in making strategically correct business decisions. “An e-commerce company, with just a fairly small AI budget can use it to find out if they should charge for shipping or if they would benefit by having free shipping and free returns. AI is useful for working out stuff like that,” says Mattias.
When Sigma commit to an AI project, they do so with a project model that, among other things, is based on well-proven iterative methods that gradually dig deeper and deeper into the analysis of the data. We get on to discuss Sigma’s methods of working with AI in order to deliver the best possible solution for the customers and Mattias explains “It is important for us to understand the data; what sort of a process is it that generates the data and also to formulate the rules for how the data will be collected. You could almost say that AI projects are the same as isolated research projects. We collect information, process it and create something operational from it,” he continues.
So, is there any cooperation between the various Sigma companies? Do they have any projects where several Sigma companies work together to find solutions for a customer? “We actually cooperate quite a lot when it comes to AI,” Mattias says. He shows how Sigma IT Consulting and Sigma Connectivity worked together to find solutions for Anticimex. Their traps now use machine learning to find the best locations and optimise the effect. The collected and processed data can later be used in Anticimex’s marketing strategy to display statistics and undeniable facts, on how effective their solution is and what benefits it offers the customer.
AI brings with it many benefits and can be put to good use by Sigma’s customers. Most things today are digitalised and, in the digital world, AI and machine learning can make a big difference.
The greatest benefit lies in the fact that we can minimise or completely remove variations in production; based on the facts available, companies can make more accurate decisions. It also creates opportunities for brand new products, business models and offers. Now that we are able to calculate and simulate many possible scenarios, AI can provide guarantees for projects and also support when planning for future investments.
There’s no doubt about it that AI, as a technique, has a bright future and we round off our conversation by talking about how Sigma plan to incorporate it into their work in the future. It’s a technique, which according to Mattias, is not particularly new, but since it is now more accessible to more people, there has been an explosion in the areas in which it can be used. “Using the exclusion method is nothing new. Sherlock Holmes is not a new character either but, with his mathematical brain, he was a master at this,” Mattias says when discussing the method of using the logic of mathematics. Sigma are committed to delivering high-quality AI solutions to its customers and plan to do so with unchanged ambition and drive. In fact it’s probably true to say that they are pulling out all the stops. “Our goal on the part of the Sigma Group is that this initiative will result in our own dedicated research on AI and machine learning in the long term,” Mattias concludes before we thank him for his time.
There is no doubt that Sigma takes this technology seriously. They have been quick to see the benefits for their customers and have realised that this area has almost unlimited potential. So, if you want to work for a company that is really passionate about this, it is Sigma you should contact. You don’t need to be a Sherlock to arrive at that conclusion.
If you want to learn more about Sigma Group’s offer within the AI area, contact:
David Österlindh, Senior Director IoT & AI
Phone: +46 733 51 48 24