Onboarding machines with just one click. How TomorrowThings is changing the IOT game with blueprints.

Oliver, thank you for being with us today and sharing your journey. First and foremost, can you give me an overview of your latest startup TomorrowThings, and the problem it is solving for your customers?

Thanks for having me. TomorrowThings develops and operates an intelligent automation platform that helps non-technical people to onboard machines with the click of a mouse. Connecting installed assets on the shop floor to the industrial internet is the very first step. Clearly, there is optimization potential towards efficiency and sustainability but it requires a digital twin first – usually, this is the process that drives up costs as it requires system integrators and domain experts, which are cost-intensive resources. What we do with TomorrowThings is give customers the technology to really automate this approach of connecting machines to enable third-party AI services on top of digitalized assets which make the world more sustainable and efficient. I mean energy costs are increasing rapidly and there is a shortage in raw materials to which machine operators must react.

What inspired you to start this company together with your three co-founders?

We have been in this business for quite some time. This is our third startup and the previous two were focusing very much on a downstream machine learning approach. What we did was using machine learning to optimize commercial buildings, as well as industrial assets from an efficiency point of view. The pain on our side was always the onboarding… “Okay, now there’s this new commercial building or this interesting machine we want to take for our journey, but the digital twin always has to come first, right?” – therefore we had to spend our own time doing this onboarding task manually. We thought to ourselves that there must be a better approach… and this was the idea for inventing a technology called blueprints that does this job in an automated way.

So let’s dive into blueprints. How do they work? How can they be used for different machines and different languages in the global market?

The blueprint approach itself is horizontal, it applies to various kinds of machines, but the first step is creating the so-called master blueprint. The master template comprises of the set of meaningful data points from which you can derive other blueprints for unknown machines in the same vertical but from other manufacturers, for example. So, by picking and applying blueprints from our library you can digitalize installed machines - one after the other. At that point, anyone can do it, you choose the appropriate blueprint to generate a digital twin for any kind of machine type.

Do you think your solution makes TomorrowThings unique compared to existing alternatives?

The blueprint approach is unique. It is very common to use self-service open-source technologies, but the issue with those tools is that they require people who are skilled in a particular domain and who can select or pick the interesting data points. The process is costly, hard, and very long. We replace this manual approach with blueprints, which are quite new to the market. There's no other company offering an automated approach to onboarding machines with just one click. It saves our customers up to 90% of their associated costs.

You recently secured one and a half million euros in funding! What are the goals you want to reach with this round of funding?

We want to productize our solution, take it to the next level, and roll it out to certain industries. We also want to scale with channel partners who could then leverage our tool chain and drive marketing and sales activities within additional verticals. We already achieved post-MVP state and are currently launching our product. Now it needs to be further automated so that we can hand over this application to end customers. If you have a machine and none of the blueprints are appropriate, you can use our AI engine to pick the relevant data points on your behalf from the interface. And if you want to store it as a new blueprint, then you can reuse this template to onboard the next machine of the same type. Right now, it’s about scaling up and setting up operations to support clients primarily in Europe and the US.

IoT technology has been around for quite some time. Why are you bringing your solution to the market now?

There are a lot of things going on the AI side with ChatGPT etc. So there's a new IoT wave resulting from the AI hype. I mean all those intelligent approaches require data. Not just data, but good quality data, and that's what we can support with our technology. Because we are experienced with machine learning ourselves, we know what is required to make AI fly. If you put bad data quality into such a model it will behave accordingly – the better the data quality, the better the results like more cost savings or less CO2 emissions. What we are now focusing on is making sure that for the digital twin, you get the highest possible data quality so it can be then consumed by any kind of intelligent model . And there's a need in the market. I mean lots of AI startups showing up right now and they require digitalized machines to jumpstart.

From an ecosystem point of view, we are becoming a sales and integration accelerator. So we're talking to machine manufacturers and machine operators and system integrators. Also, there are startups and industrial software solutions like MES or ERP and all of them are dealing with the machine onboardings. This now comes in an automated fashion and without front-up investments.

So TomorrowThings operates on a subscription basis?

Correct, it’s a subscription either monthly or upfront for a year per connected machine or connected device. You can choose from three plans. There is the starter plan with a price per connected machine and then you have the enterprise and scale plan which include a certain amount of digitalized assets. But yeah that's how it works. Building digital twins is simplified, as well as saving time and money. Just by applying the appropriate blueprint, it takes a matter of minutes which is why our approach is up to 90% cheaper compared to conventional approaches.

The overall idea is to become an operating system for machines on the Internet. What we can offer to make this happen is let's say an AppStore for easy-to-use machine blueprints, plus we support third-party applications on top of digitalized assets. In order to get there, we do offer an attractive pricing also for customers connecting hundreds of machines.

So then there would be a wide range of partners that could contribute to the open software?

Exactly. In addition to us, machine manufacturers or system integrators are allowed to put blueprints into our library. Those stakeholders have insights into a huge base of installed assets. If they provide this knowledge in terms if blueprints they can get a revenue share once their blueprints are used to create twins.  

From your experience selling to the building industry, what insights do you have that helped you build TomorrowThings?

We are experienced in the IoT and AI business. That's what we did for the last 10 years, and we know what it means to digitalize assets first and the pain associated with this task and potential data quality issues resulting from it. We did it ourselves for quite some time, then we tried to optimize and automate this process because we know there are so many companies that would benefit from cutting setup costs and connecting assets in a one-click fashion.

In the dynamic startup ecosystem, setbacks are no strangers for founders. What have you learned from past mistakes?

Making mistakes is a natural thing, right?  I mean, especially from my previous startups, learning has come from experience with scalability as well as customers’ willingness to pay for services. Enabling assets was always a painful job to be done and usually nobody wants to pay for it. It was super time-consuming and cost intensive so not supporting scalability. Let’s say I want to optimize 10 buildings in the next week then obviously you have to have more resources because people are looking into each interface and trying them to set up. As I said, it always starts with enabling interfaces first so that you can apply a machine-learning approach. We are now moving on from a more project-related business to a product-driven approach that is affordable for the industry.

Are there any trends or changes that you foresee in the industry you're working in with TomorrowThings and how you are preparing for them?

One of the major changes is actually that everything is going into cloud infrastructures, whereas before everyone was focusing on private data centers. Now, you have to convince customers (particularly German ones) that your solution is secure, that you set up a private VPN and a LTE connection for example, and that there are no negative side effects from being connected to the internet. This is one of the major concerns that our customers have – my machine is being connected, can anyone else access my assets? You need answers and technical solutions to convince them that potential negative scenarios won’t happen. The main pain point we tackle is cutting onboarding costs and enable machines for optimizations. Let's start with one two three machines, see if it’s beneficial and then we can discuss a rollout. That huge setup costs no longer exists.

As you mentioned earlier, energy costs are rising, and also raw materials are in short supply. There are also difficult geopolitical dynamics between Russia, China, the US, and Germany… How do you see this situation impacting TomorrowThings?

The increasing raw material costs or energy prices plus shortage of skilled workers are pushing towards automated solutions like ours. This is helpful for our proposition. Particularly a shortage in raw materials as well as increasing energy costs are resulting from geopolitical dynamics. Our customers must deal with it. But we can support from an efficiency point of view, for example, as we are providing insightful data streams that can be analyzed to reduce scrap or raw material waste in production.

Looking ahead, what is the overall vision and the future that you see for TomorrowThings?

We want to become the operating system for global machinery so we can create our app store with plenty of blueprints that can be used to onboard almost every machine. In addition, lots and lots of startups that already are digitalized want to optimize those assets. Our goal is to bring together customers and startups over this app store for the connected machinery. It's not an easy task for sure, but we are already in good shape, I mean we are being approached by both players already. Now, we have to make it happen.

Who are your current customers? What kind of Industries are they operating in?

Right now it’s mostly companies operating in the plastics and metalworking industry. So injection molding, CNC machines, etc. Together with channel partners we want to dig into additional industries by leveraging their domain knowledge and our horizontal blueprint technology. In addition, we are talking to device OEMs who sell IOT gateways used to connect and onboard machines. Together, we can bundle hardware with our software agent to connect machines with just one click.

Thank you very much for your time Oliver and good luck on your journey!

Thank you!