Interview Of The Week

Interview Of The Week: Dirk Slama, Bosch

Dirk Slama is a Vice President at Robert Bosch and a research fellow at Ferdinand-Steinbeis-Institute. As conference chair of Bosch ConnectedWorld,  Slama helps shape the AIoT (artificial intelligence + Internet of Things) strategy of Bosch. As Editor-in-Chief of the recently published AIoT Playbook  and Chairman of the AIoT User Group, he brings together practitioners from different industries to help advancing the adoption of AIoT. As former chairman of the Industrial Internet Consortium, he helped to create the leading ecosystem for the Industrial Internet.

Slama has 25 years of experience in very large-scale IT projects, distributed systems and IoT. His international work experience includes projects for Audi, Daimler, Lufthansa, Boeing, AT&T, NTT DoCoMo and others. He is the co-author of four books and holds an MBA from IMD Lausanne as well as a PhD in Information Systems and a Diploma (MSc equivalent) in Computer Science from TU Berlin. Slama recently spoke to The Innovator about the cultural and technical challenges involved with combining artificial intelligence with the Internet of Things.

Q: Why is there a shift to AIoT technology?

DS: A lot of companies who make physical products and equipment have adopted IoT technology over the last ten years and they have gotten significant gains from this. Many of the applications are on a level of remote condition monitoring that give them real time insights of the current state of physical products and assets in the field. This is extremely valuable, but it is only utilizing current technology to a certain point. We are now starting to see companies in the business of physical products and assets using AI in combination with IoT. The next leap is towards much more advanced applications, using AI either close to the physical product or in the Cloud.

Q: What sorts of new business models does AIoT enable?

DS: Generally speaking, there are two types of intelligence empowered by AI:  product intelligence, which involves making an individual product like a robot vacuum cleaner, more intelligent. The other type is AI in the context of swarm intelligence, which involves applying AI to a fleet of physical products or assets. This involves having thousands or even millions of products in the field and learning from all of them. For example, a smart cooking appliance connected to the Cloud would help you to collect data about how people are cooking, derive knowledge from this and recommend to a particular person a new recipe based on their individual preferences.

Q. What are the most difficult challenges for companies when scaling AIoT solutions? 

DS:We have awhole dedicated chapter on this in the AIoT playbook. It starts with really understanding the strategy requirements that you need to combine: the right product strategy, the right go-to-market strategy and the right revenue strategy. For the go-to-market strategy you need to make sure you can use the intelligence you get from the customers and translate that into customer loyalty and lead generation. For revenue generation there are different types of monetization strategies. If, for example, you use a freemium strategy at what point can you convert this into paying customers?

Q: Can you give a concrete example?

A: Take the example of AI-enabled electric toothbrushes, which are optimizing how customers are cleaning their teeth.. There are technical challenges and lots of decisions that need to be made. What changes are needed to the physical product? Do you need a smart phone in addition to the toothbrush to use the AI-enabled services? How do you get the data? Who do you market this to? Do you charge a premium for the hardware, or do you sell this as a freemium product and offer ten uses for free and if you want to use it for a 11th time you have to have a subscription plan? How do you go to market with this initially and get the early adopters? And then the big question: how do you cross the chasm? How do you go from a product designed for a smaller, highly tech savvy user base who wants to stand in front of the mirror brushing their teeth with three gadgets in their hands to something that the majority of people will embrace? How do you manage the transition in your product development? In the beginning you will need creative hackers but at some point, as the product matures, you will need to make incremental changes, and this requires different skills. You need a team that knows how to listen to your customer and utilize AIoT to understand how the customer is using the product and then use this information to continuously optimize the product by adding new software features and retraining your AI models to optimize the tooth brushing recommendations. This way of continuously optimizing your product -even after the customer has made the initial purchase -in not in the DNA of most manufacturing people.

Q: What sort of organizational change does this continuous optimization approach require?

DS: There is a whole cultural transformation that goes along with the shift to AIoT. If you are a manufacturing company and you have been doing this for decades, if not longer, then you have certain processes and you have a risk culture that is very long-term oriented. If you bring in people who are risk takers and combine them with the existing staff it is a huge challenge. We have seen digital companies that are moving into smart home with physical products experience similar learning curves. With IoT I used to talk about a clash of two worlds: manufacturing and the Internet/Cloud. Now we are adding AI, which has a culture of its own, and we have a clash of three worlds. We really need to get these three cultures working under one umbrella.

Q: Are there companies that are doing this successfully? Can you point to some?

DS: There are very successful lighthouses out there. Tesla is one of them. So are companies making advanced robot vacuum cleaners and those making smart home and lighting products. These are already multi-billion dollars markets, but it is early days. It feels like the first five years of the Internet, which is now 25 years old, so there are 20 years of innovation to come. There is huge potential.

Q: What advice do you have for companies who want to embrace AIoT?

DS: You need to combine manufacturing, software and AI mindsets and you need to get the cultural level, organizational level and the methodology right. It really starts with the business model. Do you go from the physical product to the business model, or do you take a greenfield approach and disrupt a sector with a holistic solution? Do you go in the direction of co-creation and have a partner develop the digital experience for your physical product? These are huge decisions, and you need to figure out where to start. If you start with physical product, the advantage is you know the domain inside out but the disadvantage –is you are likely to be stuck on the current focus of value creation and might not see other opportunities. If you start with a greenfield solution, with people who don’t know the domain, they are more likely to take a holistic and creative approach, but their lack of domain expertise could mean they are naïve about what can be achieved.  Once you have a business model you have to get the UX right and laser focus on this the same way that startups do – except that when you are dealing with products in the physical world, it is hard to validate certain things. If you are a software startup it is easy to put your app in the Cloud and have customers test the protypes online. If you have a digital/physical product, you need to figure out how to get ten customers to test your products in a physical, safe kind of setting, using the actual physical product. The last thing is, after the initial feedback, how do you build up an organization that constantly innovates in this combination of digital and physical settings?  AIoT requires constant innovation of software and AI models but also innovation of the physical products. How often do you update the physical product? If you don’t take a holistic approach to the software, the AI and the hardware you will fail to scale.

This Interview Of The Week article is content that would normally only be available to subscribers. Sign up for a four-week free trial to see what you have been missing

About the author

Jennifer L. Schenker

Jennifer L. Schenker, an award-winning journalist, has been covering the global tech industry from Europe since 1985, working full-time, at various points in her career for the Wall Street Journal Europe, Time Magazine, International Herald Tribune, Red Herring and BusinessWeek. She is currently the editor-in-chief of The Innovator, an English-language global publication about the digital transformation of business. Jennifer was voted one of the 50 most inspiring women in technology in Europe in 2015 and 2016 and was named by Forbes Magazine in 2018 as one of the 30 women leaders disrupting tech in France. She has been a World Economic Forum Tech Pioneers judge for 20 years. She lives in Paris and has dual U.S. and French citizenship.