Developing Decentralized Digital Intelligence AdptEVE To Understand Energy Systems

Published on:

Munich-based startup Freeelio is working on an AI with the ultimate goal to fully understand energy systems. We spoke to founder Sebnem Rusitschka about AdptEVE, decentralization, and the future of Blockchain in the energy industry.

Sebnem, describe Freeelio & its product AdptEVE in a few words.

At Freeelio we are developing a decentralized digital intelligence called AdptEVE and energy applications that both feed AdptEVE with appropriate data as well as use features, which are AI-aided. Ultimately, AdptEVE will learn to understand dynamics in energy systems. As a start, we deploy it in solar buildings for advice and adaptive optimization.

Developing Decentralized Digital Intelligence AdptEVE To Understand Energy SystemsWhat inspired you to create an energy sharing community? How did it all start?

I mastered in P2P network design and during the following R&D career at Siemens quickly was drawn to decentralized energy systems in 2007. “Will they be sharing energy like MP3 files?” was a question that formed and kind of represented my generation in my mind: We started using Napster in dorm rooms and now that we have jobs, apartments, houses, cars, solar rooftops – and an app – sharing comes naturally. So I kept researching all aspects of data management, data processing, and information extraction – with various R&D teams from P2P and Cloud Computing, to Machine Learning, to Smart Embedded Systems – for an entire decade.

Only last year, though, during a vacation in Goa amidst sunshine and power outages, all puzzle pieces came together. There are quite a few players in the end consumer market: from energy monitoring device vendors to energy equipment vendors, to installers, to new energy service providers – in developing countries and in my own backyard. All having their own sort-of communities and contributions towards clean, decentralized energy. The company I was going to build would help them create decentralized de facto energy sharing communities by sharing insights extracted from energy data.

The final spark of inspiration came when I reached out to Nick Gogerty of SolarCoin and Ed Hesse of GridSingularity: They were already building decentralized platforms and testing new incentives, means of monetizing decentralized energy and were very welcoming.

For users, your app is worry-free, meaning it calculates & distributes energy on its own. Where does Blockchain technology fit in?

AdptEVE learns patterns locally in a home or building, which means it will know when and how much spare resources are available: excess solar power, unused capacity in your battery, flexibility to defer heating or cooling for a few minutes of a few degrees etc. That’s all.

AdptEVE can then, however, be the adaptive logic of a smart contract for energy delivery that is on the blockchain. There are already quite a few companies building such constructs of smart contracts (Grid+ agent, PowerLedger neo-retailer, or the GridSingularity energy agent). AdptEVE then becomes an alternative logic that the self-sovereign users (or a tech-savvy new energy service provider) can deploy into these blockchain constructs. Another benefit is, since we must ensure that AdptEVE only learns from plausible, cleansed data, AdptEVE acts like an Oracle for these energy smart contracts.

The other side of the coin is that Blockchain technology helps us create a track record for AdptEVE. AI is a technology that would otherwise struggle with trust issues, especially in a conservative sector like energy. With the Blockchain, the predictions of AdptEVE and their return value, AdptEVE’s performance over time will become very transparent. You will not have to take a startup’s word for it, it’ll be engraved on the Blockchain.

Are you using your own app internally? How does that affect the viewpoints of the development team?

Well, it’s quite a bit different than if we’d developed an app for developers or smartphone users.

With AdptEVE we’re very conscious that end users want actionable insights. They have very low frustration tolerance with an AI-aided app that still learns. First and foremost, they are absolutely not interested in our nerdy talk. To those last two points, we as computer scientists are really blind to. That’s why user experience design trumps all at the end. Technology – AI, Blockchain – must vanish into the background. Our early adopters will help us get there quicker than we could develop in isolation.

One of the early benefits of having AdptEVE is that it can help the end user get a sustainable energy balanced lifestyle. I believe we are affected even faster than our early adopters in that regard. However, to bring AdptEVE to mainstream use, we know it must be even more carefree to use, which will hopefully be the case in 18 months to two years time.

The energy community in Europe is constantly growing, pushing more competitors onto the market. What makes you stand out of the crowd? What is the key differentiator?

There are many differentiators: We do not develop our own energy monitoring hardware or platform. We collaborate with all whose business model fosters energy efficiency and decentralized energy in developed and the developing world.

Our long-term aim is to develop an AI that understands energy systems.

For that, we would need a considerable sized data science team trained in the energy domain. Instead, we are building on our network of world class data scientists and their friends and teams who want to train the newest models and need the right representation of benchmark data and challenge from us.

Hence, the key differentiator is that we embrace decentralization in its entirety – including the financing and business models that push the “race to zero” in the energy domain. Freeelio stands for “free electricity I/O”.

What is the most memorable moment throughout the history of Freeelio?

The most memorable of all still is that moment of realization in Goa. Before I went there for vacation, my research more or less had come to a conclusion that no matter how promising those algorithms are that they would never really “know” what a power system failure is because there are rarely faults in the German power system! That seemed like a very frustrating end to a very promising research.

And yet there it was, on the most beautiful beach side ever, under the bright sun, the regular disturbing smell and sound of the backup diesel generator of the resort made up of cozy huts – whilst on the other side of the road people lived in huts with no electricity at all: the need for – and the absolute feasibility – of ad-hoc self-organizing decentralized energy exchange networks, which can be set up in a do-it-yourself manner. With all the various energy phenomena engraved in the data to let the algorithms learn from and become better at operating these networks.

When the day-to-day business of bootstrapping Freeelio wants to tear us in all sorts of directions, that memory is like a guiding star. We know we need to be practical to start a profitable business in our home market, but it needs to be as easy-to-use and deploy as if we were shipping the next release to energy monitors installed in huts in Goa. Because that’s where AdptEVE is going and beyond, the moment we are ready to scale.

This mindset has won us the right partners and an Award at the World Climate Summit COP22.

What is the biggest challenge that the company has faced?

Again, since we are a very young venture and engaging in very mindful business relationships we have not yet had a crisis type of challenge. The focus is the challenge that all new startups struggle with. In our multidisciplinary efforts as a deep tech company in a conservative sector like energy, the struggle with focus is a lot tougher.

For example, we have enough on our plate only with developing the AI, which we can focus on in liberalized energy markets or rural areas. However, since the return on our efforts will be multiplied when there is a functioning Blockchain infrastructure in the energy market – globally, there are a few fundamental enablers we need to position now. Such that when technologies have developed the required level of maturity, user acceptance comes naturally and regulators are ready to support the innovation.

If there is one thing you could wish for in improving the startup ecosystem in Europe, what would it be?

Regulatory sandboxes in Europe are of utmost importance. We need to face the reality that there is no European single market anytime soon when it comes to data economy, digitalization – or not to mention the token economy. Whilst it is already hard to find a foothold market and solution to all these potentials in one country if a European startup wants to scale, it has only one next market to go: the US. Not only due to the sheer size of the US market – but also the “permissionless innovation” culture.

The European market is big enough, many deep tech innovations are springing in Europe – so all we are missing now is a more welcoming approach to technological experimentation: If the existing rules and regulations do not apply to the new possibilities, regulatory sandboxes can help startups, users, and regulators to learn how to best enable technological advancement whilst protecting end users – as well as European innovative capital.

According to the credo: “Be the change that you wish to see” in the startup ecosystem, Freeelio became a founding member of the German Blockchain Association, in which I lead the Energy Working Group. We have joined efforts with many Blockchain startups and experts who want to see Blockchain as a fundamental infrastructure of our digitalizing societies. We are also reaching out to like-minded trail blazers in Europe to create a safe space for digital innovation in the traditional sectors like Energy and the Public Sector.

What piece of advice would you give fellow founders for their startup?

That would be an adaptation of Steve Blank’s “The minute you take money from someone else, their business model becomes your business model”: A startup’s sole purpose is to find product/market fit for their idea or “vision”. If during that time you mistake your market as the VCs and your pitch as your product (which happens gradually from incorporating one VC feedback to the next), you will have built a company that runs from fundraising to fundraising until it runs out – and the sorry bottom line is you’ve spent 5 years of your life doing just that.

If you are building a product which creates network effects and thrives on them – then you now have a new tool in your financing and business model belt: the Token Model. Research it, follow it through 2017. This space will mature in the next two years. The token model aligns incentives of users, developers, investors, and other contributors – and it is the ultimate customer validation effort.


For more on innovation in the energy sector read StartUs Insights’ Breakdown on Startup-Driven Innovation in the Energy Industry.


Sharing is caring!