“With smart enterprise, we’re placing our focus on human beings rather than on AI technology”
Dr. Andreas Meier did his doctorate in the field of artificial intelligence (AI) at Volkswagen and spearheads the team promoting the use of AI in knowledge work
Mr. Meier, what does “smart enterprise” actually mean?
Roughly speaking, we use the term to describe the strategy of how we want to use artificial intelligence for knowledge work within the Group.
That immediately sets off alarm bells for employees. Does it also mean how artificial intelligence can replace our jobs?
Not at all! Rather, we ask ourselves how artificial intelligence can help us with knowledge and office work within the Group. We believe that artificial intelligence doesn’t take away our jobs. On the contrary, it helps employees and will lead to new types of jobs. That’s because artificial intelligence has different strengths than human intelligence. We want to combine both in the best possible way. But at the moment we’re very technology-focused in the Group and sometimes lose sight of the user. With smart enterprise, we’re placing our focus on human beings rather than on AI technology and at the same time want to see what opportunities open up for new areas of activity.
Let’s speak in concrete terms. What are the objectives of smart enterprise?
We have three. First, we want to speed up office and knowledge work. Think of the reporting apparatus, for example. There’s a lot that can be automated there, for instance with Robotic Process Automation (RPA) technology. Second, we want to liberate employees from repetitive tasks so that they can focus on important, value-creating things. Chatbots are good for this, as they allow users to clarify certain questions in a natural interaction with a program rather than with a person. Third, we want to make data more useable for our employees. In many cases they have a real flood of data at their disposal. But until now they have not had the possibility to get the exact information – or pieces of related information – that they specifically need for their job. We want to change that.
So that means that smart enterprise is focused on office and knowledge workers rather than on production staff?
That’s right. We are primarily focused on the many activities on the administrative side of things, such as human resources, finance, procurement, quality assurance, etc. Depending on the issue at hand, however, it can also have a connection to logistics, quality control or even production. But traditional production is not currently what we are working on.
How long has your project team been in existence, and what are your backgrounds?
Since October 2016. We’re a very diverse team with IT people, business people, HR people and strategists. But we also need all of them to get this wide-reaching strategy up and running.
Mr. Meier, how do you approach concrete projects?
When we speak with departments, the first thing we ask is: What problems do you have? How can we help you? Then they explain their work processes. Our aim is to find everything that can quickly help office employees with their work. We want to free them of the dull tasks. The colleagues with whom I’ve spoken so far have never been skeptical but rather are always very interested in how we can help them with their work.
How exactly does such support work?
Take Idea Management, for example: in that department they’re tasked with evaluating the ideas of employees, obtaining expert opinions, submitting financial and feasibility forecasts. At present, a lot of this work is done manually. For each new proposal, an employee has to set the whole process chain in motion again and then work through it. We’ve now developed software for this, which is currently still in testing. It encompasses the enormous dataset of all proposals submitted so far. At the push of a button, the employee now sees whether there have been similar proposals before and, if so, who did the assessment. The search engine we developed runs in real time and suggests topics while the employee is still typing in the request.
What else can the software do?
Intuitively filter ideas. For example, employees in Idea Management can ask the program: “Show me all the unrealized ideas that are more than three years old” or “I would like to see all ideas on the subject of Volkswagen currywurst or lubricating oils.”
What kind of savings does the software achieve?
For starters, it saves a lot of time previously spent on tedious, manual tasks. In that sense it also benefits the Volkswagen Group financially, as the employees in Idea Management can then use that time on other matters. My personal vision is that this will free up resources to better serve the people submitting the ideas. Here are a few figures: Idea Management at Volkswagen AG has existed for 68 years. Last year it generated savings of 124 million euros and rewarded employees for their commitment to the tune of 22 million euros. Given that there were almost 55,000 ideas submitted in 2017 alone, you can imagine the huge volumes of data that can be searched in a matter of seconds. We’re talking about over 1.36 million ideas!
The software was created internally?
Yes indeed. We work with the systems already in place and must therefore ensure that the new software works well with them too. We canʼt simply buy standard software on the market, as a program like ours that has all these functions and is compatible with our systems is simply not available to buy anywhere. So we developed it ourselves and ran various tests to get the necessary approvals. And lastly, we also created the user and developer manuals. All of that happened within one year.
Are there any further concrete examples you can give of the smart enterprise strategy?
There are many, and I am really delighted that so many different departments, including Finance, Procurement and IT, are vigorously implementing projects. Our colleagues from IT, for example, are currently developing a chatbot that can help users with their IT problems. As a user, I communicate with a system like I would do on WhatsApp, describe my IT problem, and the bot then tells me how I can probably solve it. So the employee gets fast assistance rather than sitting there waiting on hold on the telephone, which brings tangible relief to everyone involved.
Of the Group’s 642,000 employees, 308,000 of them work in the administrative area, in other words not in production. There are tens of thousands of processes being carried out daily in the offices that could certainly be made more efficient. So, are you and your team set for life in terms of workload?
Looking at the numbers alone, that would certainly be the case. But over the medium term, we want to enable the employees to use artificial intelligence and related technologies themselves to make their own processes more efficient. For only they have the expertise in their particular job and understand what processes come before and after. We provide the technological approaches, then together with our colleagues in IT use the employees’ expertise to implement them. It’s very important to me that colleagues learn more about the technology and eventually become able to use it on their own. Only through this sort of self-help will we ultimately be able to achieve high scalability within the Group.
But then at least one job would be lost – yours.
(Laughs) Although it might sound strange, I actually do want to make myself superfluous in this context. After all, there are plenty of other exciting issues to tackle. I would like to enable the departments to independently create their own artificial intelligence together with the IT experts. The combination of people and machines is unbeatable.