17-18 June 2025 - Maison de la Poste, Brussels, Belgium
When technology is applied to areas where important societal challenges arise, it can help solve problems that would otherwise be impossible to address. With the support of IoT and AI, predictive maintenance on large infrastructures, such as bridges, can prevent significant issues, from minor traffic disruptions to catastrophic collapses. Theodoor Van Der Klaauw shares how his research on developing predictive maintenance solutions for the Zeeland Bridge combines IT and AI to serve both people and the environment.
The integration of IT and AI signifies a paradigm shift towards intelligent asset management, reducing downtime and maintenance costs while improving safety. "Digital innovations can lead to sustainable infrastructure management, enhancing quality of life through social, economic, and environmental benefits," Theodoor Van Der Klaauw explains, highlighting how these technologies are transforming infrastructure operations.
This article explores predictive maintenance’s comprehensive impacts, illustrating how IoT and AI advancements are practically implemented and further developed. Using Zeeland Bridge as a case study, insights from Theodoor Van Der Klaauw’s research highlight technology's critical role in maintaining public assets at scale.
Spanning the Oosterschelde estuary, the Zeeland Bridge connects several key regions in the Netherlands. With 54 striking concrete pillars spread over 5km, it is essential infrastructure, supporting daily transit for over 10,000 vehicles. "The bridge is indispensable, not just for its monumental status, but for regional economies and community life," Van Der Klaauw notes.
Additionally, the bridge significantly impacts maritime traffic, opening roughly 4,000 times annually for around 20,000 ships. Van Der Klaauw emphasises this complexity: "Maintaining multimodal functionality requires meticulous monitoring to ensure operational continuity, minimising disruptions and sustaining economic stability."
Constructed to enhance connectivity, the Zeeland Bridge facilitates commuting, emergency services, and regional supply chains. According to Van Der Klaauw, "Predictive maintenance ensures reliability, influencing economic productivity and social welfare profoundly."
Theodoor Van Der Klaauw’s research journey began academically in Rotterdam, deepening through Oxford University’s Sustainable Urban Development post-graduate. He recounts: "My studies combined sustainability principles with urban technology, aligning with global infrastructure practices, especially during the disruption caused by COVID-19."
When choosing a research topic, Van Der Klaauw initially reached out to the Dutch Ministry of Infrastructure. "At a national infrastructure event, I met the (former) Minister who advised me to talk to a colleague at Rijkswaterstaat, the operational arm of the Dutch Ministry of Infrastructure," he recalls. These conversations quickly highlighted the potential of sustainability through data-driven asset management, and the 'ISI'-framework (Schinkel et. al, 2023) was selected to study economic, social, and environmental impacts.
"After several discussions," explains Van Der Klaauw, "we realised that embedding sustainability into asset management was essential." This approach allowed him to merge theoretical concepts with practical application. To gain tangible insights, he aimed for a case study where his background as a consultant could be effectively applied, ensuring the research would be both relevant and actionable.
Ultimately, this practical perspective led Theodoor Van Der Klaauw to select a particularly advanced use case in the Netherlands: the Zeeland Bridge. "We decided to focus on an innovation lab called CAMINO that was already experimenting sophisticated asset management techniques," he notes. The Zeeland Bridge provided a concrete opportunity to explore sustainability practices deeply within real-world infrastructure management.
Two significant challenges emerged from Van Der Klaauw’s research. "Firstly, the bridge's continuous availability was critical. Disruptions affect everything from commuting to logistics, demanding predictive strategies to maintain service reliability," he states.
Secondly, data integration posed a considerable challenge. Van Der Klaauw highlights: "Despite sensor technology monitoring bridge vibrations, standardised and calibrated data was lacking, impeding accurate predictive analyses. Without data standardisation, predictive maintenance remains challenging."
Van Der Klaauw highlights that inconsistent data significantly influenced his research approach. "Because of that lack of data consistency, the results of my study are based on expert knowledge rather than infrastructure data," he explains. This limitation necessitated a methodological shift, relying heavily on qualitative rather than insights derived from asset data. Quantitative results were collected from expert surveys.
To address this challenge, Van Der Klaauw engaged broadly with stakeholders across sectors. "I've interviewed public infrastructure experts, service providers, and private sector experts," he notes, emphasising the diversity of perspectives.
When it comes to solutions, Van Der Klaauw focuses on complete data systems: "Effective integration of diverse data sources is essential. It requires organisational shifts and technological innovations to ensure consistent, reliable predictive maintenance outcomes."
“To fully harness the potential of predictive maintenance,” he concludes, “it’s essential to harmonise asset data, making it suitable and effective for large-scale analysis.”
Once the information is standardised, the Zeeland Bridge team can collect input from the sensors and learn from these. “And that's where AI comes into play. AI learns what a normal state looks like” details Theodoor Van Der Klaauw. “And then you can use AI to identify, to build those patterns, to understand what a healthy performance looks like, and then to identify what the differences are, which give you a picture of the deviation between the expected performance and the real performance.”
AI-driven systems monitor structural data continuously, rapidly identifying performance deviations. "Early anomaly detection allows preemptive interventions, reducing unexpected failures and downtime," Van Der Klaauw explains, demonstrating predictive maintenance's effectiveness.
Furthermore, AI has proven to be extremely valuable for optimisation purposes in asset lifecycle management. According to Van Der Klaauw, "AI can help to prioritise asset inspection, maintenance and replacement activities based on parameters such as risk, resource availability, budget availability, and carbon impact, enhancing efficiency and conserving resources, thereby improving long-term asset management."
Van Der Klaauw strongly advocates for human-centred technology solutions. He emphasises: "Predictive maintenance addresses contemporary challenges like ageing infrastructure and skill shortages, offering safer, more efficient alternatives to traditional, resource-intensive methods."
Environmental sustainability also plays a pivotal role. Van Der Klaauw argues: "Predictive maintenance extends infrastructure lifespans, reducing the environmental impact compared to constructing new facilities. Building new bridges involves significant emissions and resource consumption through scarce, expensive materials."
Integrating global environmental costs, including digital technologies and AI, into infrastructure decisions is crucial, according to him: "Predictive maintenance provides financial savings and reduces ecological footprints, making it a prudent environmental choice and contributing significantly to Sustainable Development Goals (SDGs) 8, 9, 11, and 13."
A holistic perspective allows us to clearly see how digital technologies can reduce the overall carbon emissions throughout an infrastructure lifecycle, while also considering the emissions generated by these technologies themselves. Beyond environmental and operational considerations, Van Der Klaauw draws attention to the reputational risks involved in infrastructure failures.
"One critical point that emerged during my research is that I think the social damage and the reputation damage for these organisations caused by infrastructure failure, is arguably more important, than the negative environmental impact of training models and deploying IT solutions," he stresses. In this context, the strategic value of predictive maintenance extends far beyond efficiency: it plays a critical role in preserving public trust and institutional credibility.
Predictive maintenance, driven by AI and IoT, marks a transformative advancement in infrastructure management. Theodoor Van Der Klaauw’s Zeeland Bridge case study vividly demonstrates technology’s capacity to enhance asset reliability, safety, and sustainability. He concludes: "A holistic approach ensures enduring social, economical, and environmental benefits, aligning infrastructure management with global sustainability targets."
Theodoor Van der Klaauw is Managing Consultant at IBM. At GreenTech Forum Brussels 2025, he is speaker. He will lead the conference: “Digital Twins: a way for organisations to take better decisions for ecological transition?”.
Disclaimers: This study has been conducted at personal title by Theodoor Van Der Klaauw since IBM has not provided any services nor technology to this project. All statements and thoughts are those of Theodoor Van Der Klaauw, not those of IBM.
Article written by Rémy Marrone for GreenTech Forum Brussels
.png){kind=small}