Could you briefly introduce yourself and the role you have in the organisation?

My name is George Tzagkarakis, and I am a Principal Researcher at the Institute of Computer Science of the Foundation for Research and Technology-Hellas (FORTH-ICS), in Heraklion, Crete. FORTH-ICS is one of the largest and most established research institutions in Greece and South-Eastern Europe, with a long track record in data science, AI and networked systems. I am based in its Signal Processing Laboratory.

I hold two PhDs, one in Computer Science with a focus on statistical signal processing and one in Finance with a focus on risk quantification, and I have worked for over fifteen years across signal and data processing, machine learning and computational finance, in both academic and industry settings.

I lead the lab’s activities on Decision Intelligence Systems, and I coordinate the TwinODIS project, setting the scientific direction and making sure we deliver on the objectives.

Could you explain your projects and results you brought to the widerAdvance Facility in simple terms? What problem are you trying to solve?

The project behind our participation is TwinODIS, a Horizon Europe Twinning project we run at FORTH-ICS with Université Paris Cité and Erasmus University Rotterdam. The aim is to strengthen FORTH’s capacity in decision intelligence and make the institute a reference hub for data-driven decision support in Greece and the wider region.

Smart water is one of our application domains, and the result we brought here is FORTH SmartWater. Water utilities lose a lot of their water to pipe leaks, usually working with ageing infrastructure and tight budgets. Large operators can afford advanced monitoring platforms, smaller and mid-sized ones cannot, so leaks and quality problems go unnoticed for too long.

FORTH SmartWater sits on top of a utility’s existing SCADA system, so there is no hardware to replace. It takes the sensor data they already collect, flow, pressure and quality, and uses AI to spot leaks, forecast issues such as pressure drops or equipment failures, and monitor water quality in near real-time. Operators get an alert within minutes rather than days. It is already deployed at TRL 9 and was built around what the operators told us they actually needed.

What were the biggest challenges in turning your research results into something that could be used beyond the project?

The first was the gap between an algorithm that works on clean data and a tool an operator can use every day. Real utility data is messy, every SCADA setup is different, and what a researcher finds interesting is not always what an operator needs at seven in the morning. We only got past that by working closely with water utilities and rebuilding parts of the tool around their feedback.

The second is the customer base. Municipal utilities have very tight budgets and slow public procurement, so the licence had to be initially priced low enough to be adopted at all, which in turn puts real pressure on the business model.

The third, and the biggest, is that we are a research lab, not a company. We protected the software under copyright, but we have no freedom-to-operate analysis, no patent strategy and no business plan yet.

And finally, scaling out. We have traction in Greece, but no structured benchmarking against the commercial platforms, so we do not yet know how we really compare internationally.

What made you decide that this was the right moment to join the widerAdvance Facility?

The technology is ready and the business side is not. FORTH SmartWater is at TRL 9 and already deployed in real environments, so the bottleneck is no longer technical, it is commercial, and that is exactly where the Facility supports us.

The timing also fits TwinODIS. We are partway through a 36-month project, so whatever we build here we can still feed back into the project and into the other application domains, rather than receiving it once everything has closed.

What kind of services did you access in the widerAdvance Facility?

At the introductory workshop online, we consolidated our KERs and clarified the exploitation intentions table and the use model for FORTH SmartWater. That session also introduced the Value Proposition Canvas (VPC) and covered dissemination strategy.

The review workshop on site was hands-on. We worked through the VPC with the other participants and reduced the value proposition to a single AdLib sentence. The material on exploitation strategy, the Characterisation Table and the Exploitation Roadmap, gave us practical know-how we did not have.

We also used the supporting material, in particular the Lean Canvas and the guidance on freedom-to-operate and IP strategy. That last one was directly relevant, since we had copyright in place but no FTO analysis at all.

Looking back, what has been the most valuable thing you have gained from participating in the widerAdvance Facility?

The Value Proposition Canvas, and specifically being pushed to reduce it to a single AdLib sentence. It sounds like a small exercise, but it forced us to state who FORTH SmartWater is for and what pain it removes, in one line. As researchers we tend to describe what the tool does technically, which is not the same thing.

Close behind, the Exploitation Roadmap, which turned a vague intention to commercialise into concrete steps in a sequence.

The peer discussions also mattered more than I expected. Other teams were struggling with the same things, pricing, IP, being a research lab rather than a company.

How the widerAdvance Facility and its service contributes to the valorisation process of the results?

It bridges a gap that most research projects fall into. Good results are produced and then nothing happens, because nobody in the lab has the tools or the language to take them further.

Concretely, it moves you from describing a result to defining who it is for and what value it creates. The VPC does that, and the Characterisation Table and Exploitation Roadmap turn it into a plan with steps and ownership.

It also builds capacity rather than just a document. The team now knows how to use these tools, so we can apply them to our other application domains. And the FTO and IP guidance de-risks the process, because valorisation stalls quickly if the IP position is unclear.

Has your approach to engaging stakeholders, potential users or industry partners changed?

Yes. We used to lead with the technology, what the tool does, the algorithms, the performance. That is the researcher’s instinct. Now we start from the other side: what the pain is, what it costs them, and where our tool fits.

The second change is timing. We used to approach stakeholders late, with the product almost finished. Now we engage them much earlier and treat them as a source of insight rather than only as customers.

Has the widerAdvance Facility influenced the way your team works together or plans future activities?

Definitely. Exploitation is no longer one person’s job. It used to sit with me as coordinator while everyone else got on with the research. The team has now been through the same tools, so we share a vocabulary, and when a result appears the question of who it is for comes up much earlier.

We also treat the Exploitation Roadmap as a working document rather than something written for a deliverable. It gets revisited and it shapes what we prioritise.

The same thinking is now applied to the other TwinODIS domains, namely, maritime logistics and energy, asking early which results could become KERs. And it left us with a clear list of gaps, FTO analysis, business plan and benchmarking, which are now planned activities with owners.

Do you feel your organisation is now better prepared to maximise the impact of future research and innovation projects?

Absolutely. We now have the tools and, more importantly, people who have used them on a real case rather than in the abstract, and we are already applying the same thinking to our other domains.

TwinODIS reinforces this, because capacity building is what the project is for, so what we learned feeds into training and into how the institute works more broadly.

What is still missing is the purely commercial side. Turning a product into a business needs skills and structures a research lab does not naturally have. We are much clearer now about where that gap is.

What are the next steps for your project and its KERs?

First, closing the gaps we identified: a business plan for FORTH SmartWater, the freedom-to-operate analysis and a structured IP strategy.

Second, benchmarking and international positioning. We have traction in Greece, but we need to know how we compare with the commercial platforms before pushing further out. This feeds into a market intelligence framework we are building at FORTH.

Third, returning to the market with a stronger case, through matchmaking and pitching events and by re-engaging the stakeholders we spoke to before, this time with a plan behind us. In parallel we keep scaling adoption and adding features. Beyond FORTH SmartWater, we are running the same process on the KERs from the other TwinODIS domains.

Would you recommend the widerAdvance Facility to other organisations? Why?

Yes, particularly to research organisations in widening countries, because we are typically the ones with the results but without the machinery to do anything with them.

The training you get from the widerAdvance Facility is practical. It is not a lecture on why exploitation matters, you work on your own case with tools you keep using afterwards. In our case, we came out with a sharper value proposition and a roadmap.

The peer element is a real part of it, too. Being in a room with other teams facing the same problems is more useful than it sounds, and some of those contacts will last.