I’ve got a belief that the ‘winners’ in IoT aren’t necessarily going to be the traditional technology vendors. Yes there will be some that do well on the hardware side, but these will quickly become commodity products and struggle to add value down the line. Those that I think will come out on top, are the ones that are already selling physical products to customers – but choose to embed IoT across their product line and adapt the culture of their business to suit new ‘servitization’ models.
Bosch could potentially be one of those companies.
This month at MongoDB’s annual customer event in London I got the chance to sit down with Dirk Slama, director of business development at Bosch Software Innovation – Bosch’s Internet-of-Things division, which has built a technology stack and consultancy service to help its internal business units adapt to the IoT way of doing things, as well as to sell to any other company that wants to invest in IoT.
Bosch SI’s stack relies on MongoDB for the backend data collection, which is why Slama was in London at Mongo’s event. Slama said:
Over the past five years it has become clear to us that the world that in which we have been living in for the past 125 years is now really coming to a turning point. Industries have been completely transformed by the internet in the last 25 years, but that has not really happened to the manufacturers. For most manufacturers the internet has been a nice additional sales channel. But this is now changing with IoT.
This is about different or new customer value propositions by making products more intelligent, by adding local intelligence, but also by adding cloud-based intelligence. What we are doing is almost every division in Bosch now has people focused on identifying the best IoT use cases, IoT solutions.
And then also understanding not only how this changes the customer proposition, but how this changes our own value chains. How can we make use of the data in the future by being better at designing our products, marketing our products, selling our products, servicing our products? We see this has a huge impact on all of these aspects.
This is something that I’ve highlighted before as being key to transforming your business with the Internet-of-Things. It’s about shifting the culture of your company away from a one-off product proposition, where the lifecycle can be a number of years, to a position where you are selling products and you are constantly engaged with that product/customer throughout. This impacts both customer service, as well as product development, as it allows you to receive data on how the product is being used and what could be improved.
None of that is easy, especially for companies that have not been operating with a product mindset for decades. Slama said:
This is going to be different from business unit to business unit, from product to product. Because of this we set up Bosch Software Innovations, which is the unit I work for, and our job is to help Bosch customers and Bosch business units to help them successfully adopt IoT business models and to implement them using our software.
We have built out a software sack that starts with the ability to integrate sensors, capture data, pre-filter data locally close to the asset, transfer the data over a local carrier network to the backend, and then the backend captures the data from the assets and devices. The product we are using for device data capturing is MongoDB. From there it goes on to analytics and from analytics it goes on to actions
Slama explained that the use cases thus far have varied from business unit to business unit internally at Bosch, but have also varied from customer to customer. He said some of the more successful examples include:
One is called structured field data, which gets data from all kinds of components inside of cars. Whether it’s injection pumps, windshield wipers, lamps, engine data etc. This data basically is extremely valuable for our engineers who design these physical products because it gives them real-life usage data.
Another fun and interesting project was a lawn mowing robot that was IoT enabled so that you can basically easily customise it using your mobile phone to stop it working certain days and to combine this in the future with weather forecast data. In a similar area, we just won a prize for a project for deploying sensors in agriculture for asparagus. Asparagus has to be harvested at a specific time, so that when it comes out of the soil you have a window of about four hours before you need to harvest it. We are using temperature sensors to inform the farmer about the perfect time to harvest the asparagus.
We are also seeing a lot of traction in manufacturing – predictive maintenance, preventive maintenance etc. We also see a lot of traction in smart city type environments, although that is still a very ill defined concept and everybody you talk to says something different. But we have done a great project with the city of Monaco and we have done a project in San Francisco.
Slama highlighted two key challenges regarding the implementation of IoT business models. One is integrating the ‘product’ people within your business with the ‘internet’ people, making both teams work towards the same goal. Given that these two operate in completely different ways, it’s not an easy task. Slama said:
This is the biggest challenge with IoT. We have just published something ourselves where we talk about the clash of two worlds, where you have people that focus on physical products who think about development lifecycles of three to five years, who know that after the start of production there’s no stopping it and what is produced has to be deployed in the field and support a lifecycle of decades.
On the other hand you have the internet guys are thinking about minimum viable product, perpetual beta, daily updating your services. That’s a tricky value proposition, you have the physical assets out in the field and then software that’s updated daily.
If we are selling machine components to a manufacturing customer, who basically says that this is X million of investment and has to run for ten to twenty years, that’s a very mindset and different world. Bringing these two worlds together is a challenge.
So how do you overcome it? From Bosch’s experience, this requires top-down support and a complete organisational overhaul of the business to align the two strategies. Slama said:
You have to build up the right organisation structure and governance in your organisation. If I am looking at Bosch, for us, this is really a major organisational transition. It starts at the board. At the board level it is supporting this, that we have the right governing bodies in the organisation that can bring together the people from different divisions.
Or take the fact that Bosch Software Innovations was set up within the company as a catalyst, but also for our customers to help drive these changes. It’s complex and you have to address
it at all levels.
However, even if this organisational overhaul is successful, the other key challenge with shifting to an IoT business model, where servitization is the name of the game, is changing your financial and revenue expectations. As we have seen with the cloud computing vendors, this isn’t easy. Slama said:
Servitization has a lot of benefits and yes in general the argument from a financial point of view means building up a much more stable revenue stream, is a good argument.
But if you already have a going concern that’s based on up-front revenues and then you start saying that in three years from now you will have more stable revenue streams across your subscription base, but for the next three years you will have to take a cut of revenues of 80% until we have built that up. That’s a hard sale for your CFO.