“Strategy today is less about crystal balls and looking around corners and more about testing, listening to feedback, and changing fast.”

Martin Lundqvist always dreamed of starting his own company but enjoyed management consulting too much and ended up becoming a Mckinsey Digital Partner. After 17 years at the top tier firm, he realized his dream and joined the scale-up Arundo Analytics where he is the COO today.

As a former McKinsey Digital partner, can you tell me why companies find it so difficult to become digital that they pay their healthy fees?

The first reason is usually that they struggle to define what becoming more digital means — and why they need to do it! Is it to grow an existing business? Shape new business models? Become more efficient?

The second challenge is to figure out where to start. The most common mistake companies make is to have a list of 50+ initiatives that all have to happen urgently. The most successful cases are companies that are able to focus on a handful of initiatives with a deep connection to Business Operations. The problems you aim to solve must match the business perception of problems to be fixed.

Another common challenge is gathering the right competence to deliver. The best results usually come from cross-functional teams which means that you need both Tech talent and talent from the core business. The Tech talent usually rather flock to start-ups or the Googles of the world. When you ask for resources from the core operations, you are more likely to get the ones that are available, than the ones you need.

So I’d say companies like McKinsey help corporations focus on the right things, provide some hard to get talent, and the fees are motivated by what a successful outcome can give. Which is often a lot.

What does your current company, Arundo, do?

We build software products for sustainable industrial operations using advanced analytics. Our products allow reliability and maintenance engineers to spend less time chasing current problems, and more time focusing on avoiding those problems in the first place. By understanding the current equipment health and predicting anomalous equipment behaviour, we help extend equipment life time, increase productivity, reduce costs and lower environment footprint.

Arundo was born out of the founder’s realisation that industrial companies generate petabytes of data that is barely used. I’ve been involved in different capacities from the start, because I saw the need for a great solution in this space.

What is the biggest challenge in introducing new technology to heavy industry?

It is with good reasons a conservative industry. Safety, security, predictability and operating procedures are super-important. New technology challenges that status quo and risks disturbing the stability of processes and no-one really wants another piece of software to worry about.

That said, change is necessary to achieve real improvements. If you’re serving the heavy industry, you need to spend time with the end-users to make sure that your product truly adds value to their work; and that it can be tailored to integrate seamlessly with their established operating practices.

In turn, heavy industry needs to develop a culture of frequent realization. Strategy today is less about crystal balls and looking around corners and more about testing, listening to feedback, and changing fast.

You also need Rebels. People that dare to confront the established ways of doing things and have enough credibility in the organization stand up against the “tissue rejection” that new ideas bring.

Why should we trust Artificial Intelligence?

Not sure you should!

Question is, WHEN should you trust artificial intelligence and under which conditions. And here, I would argue that it is not much different from how you pick e.g., a doctor. You give it a try. You look for consistent results. You look for explainable results. Good AI (in our world) implements the simplest possible solution; has it’s foundation in first principles engineering; its results can be clearly traced back to the data input. And its output can be easily translated into action.

What is the difference between good use of Artificial Intelligence and bad use?

I personally think the purpose plays a huge role. When it comes to Safety, I think we need to place trust in the people and organizations responsible for ensuring that risks are addressed on a system level. We have to trust them to perform their work ethically and responsibly and give them the tools needed, but together with strict rules and frameworks like GDPR.

Buddywise is a start-up hell bent on saving lives and preventing injuries with computer vision!