As DARROW enters its final year, we are not only highlighting the project’s technical results, but also the people behind them. Throughout this interview series, DARROW partners reflect on the project’s outcomes, the challenges they faced, and the lessons learned along the way.
In this interview, Ruud Kassing from Haskoning reflects on what it takes to achieve secure AI integration in wastewater treatment. He shares his perspectives on the role of integration, security, and operational reality in making advanced data-driven tools usable in wastewater treatment plants.
Tell us a bit about yourself and your organization. What role have you played in DARROW?
My name is Ruud Kassing. I am the Control Software Lead at Haskoning. Our team develops reliable control and integration software for waterboards and utilities, helping them to improve their day-to-day operations by providing them with systems that are effective, safe, maintainable, and trusted by operators.
As a consortium partner in DARROW, we are responsible for delivering the integration platform in which all the software components developed by the other partners are deployed. My role has been to ensure that the overall solution is technically sound and deployable in a real operational environment. We established the connection to the treatment plant’s PLC through an integrated platform so that relevant operational data could be made available to the partners’ software modules. Because wastewater operations are increasingly being treated as critical services with high cybersecurity expectations, we also treated secure connectivity and access control as core design constraints from the outset. We supported the partners in running and monitoring their models in a controlled way, translating research needs into practical interfaces and constraints that fit plant operations.
In your view, what is the most valuable innovation or tool that DARROW has developed?
The most valuable outcome of DARROW is a practical blueprint for integrating advanced data-driven tools into the existing PLC/SCADA infrastructure of a treatment plant in a secure and effective way. The project made very clear what is required beyond algorithms: robust access to operational data, explicit interfaces, clear responsibilities, and continuous monitoring so that experiments can be run safely and transparently.
A key insight is that security and governance must be designed alongside data access and integration, especially in sectors covered by EU cybersecurity regulations like NIS2. Combining data availability, integration discipline, and operational alignment reduces risk and makes it easier to repeat similar implementations at other plants, even when the underlying models evolve over time.
If you had to describe DARROW in one sentence, what would it be?
DARROW shows how to connect advanced data-driven innovation to real treatment plant operations by aligning data, integration, and people from the initial concept stage through to monitored deployment.
What is something you have contributed to DARROW that you are especially proud of?
I am especially proud that we transformed DARROW from research conducted alongside a plant to research that can safely run within the plant. By connecting to the PLC and creating a reliable integration and monitoring setup, we have enabled our partners to work with real operational data and run their models in a way that is transparent for engineers and operators.
This groundwork, which included interfaces, operational constraints, monitoring, and secure access to operational data, may not be the most visible part of the project, but it is what makes innovation transferable and ultimately deployable in practice.
What was one of the biggest challenges you or your team faced, and how did you overcome it?
One of the biggest challenges was aligning a multidisciplinary group where research teams move fast, while operational environments demand stability, clarity, and safety. Even when working towards the same goals, expectations can diverge if the practical details of integration and deployment are not made explicit.
We addressed this by making integration the focal point. We agreed on clear interfaces to the PLC data, defined how reinforcement learning and other machine learning models would be run and monitored, and established a shared understanding of operational constraints. Since operational technology environments have stricter expectations than typical IT experimentation, we also aligned on secure ways of exposing data and running components. This approach kept progress under control and transparent and ensured that what the partners built could actually be evaluated in a real-world setting.
What advice would you give to a future project team taking on something as ambitious as DARROW?
Invest early in the operational and implementation phases, not just the innovation itself. Make data access, interfaces, responsibilities, and monitoring explicit from the start, and revisit them regularly as the project progresses.
Treat integration and deployment as core workstreams, because that’s where ambitious ideas either become trusted tools or remain prototypes. If you build a safe, transparent, and secure run-time and monitoring setup early on, you create the conditions for research to have a lasting impact.
What kind of change do you think DARROW can bring to how we manage water and resources in the future?
DARROW can help the sector transition from isolated pilots to a more consistent approach to adopting digital decision-support and optimisation tools in daily operations. By demonstrating how to connect plant data, run models in a controlled environment, and continuously monitor them, it lowers the barrier for utilities to explore advanced methods responsibly.
Over time, this approach can support more predictive, efficient plant management, using energy, chemicals, and infrastructure more effectively. It also creates a clearer pathway for scaling digital innovation across the water sector. Security-by-design will be integral to this process, since these systems are increasingly located in regulated and high-impact operational environments.
Technical terms explained
- PLC: Programmable Logic Controller; real-time control computer for plant equipment such as pumps, valves, and aeration.
- SCADA: Supervisory Control and Data Acquisition; operator monitoring and supervisory system.
Ruud highlights a central lesson from DARROW: secure AI integration requires more than strong algorithms. By focusing on integration, monitoring, and security from the outset, the project shows how advanced digital tools can move beyond pilots and become part of everyday plant operations. The experience Ruud shares here points toward a future where digital innovation in the water sector is not only ambitious, but also trustworthy, scalable, and grounded in operational practice.