TNO and Rijkswaterstaat investigate use of drones for inspection of locks, dams and weirs

Wouter Hoefnagel
Wouter Hoefnagel
23 March 2026
3 min

Locks, dams and sheet piling are crucial for the safety of the Dutch water infrastructure. However, inspecting these underwater structures is not easy. For instance, traditional diving inspections are risky and labour-intensive. TNO and Rijkswaterstaat are therefore investigating how underwater drones can make the work safer, more efficient and more accurate.

Rijkswaterstaat manages around 6,500 engineering structures, ranging from bridges and tunnels to locks and dams. Much of this infrastructure is partially or completely under water. In addition, water boards, provinces and municipalities manage a wide range of other underwater objects, which are often smaller.

'Targeted maintenance is essential'

Inspecting these structures is crucial, partly because of growing maintenance costs. For instance, the annual cost of replacing and renovating the structures is rising from around €1.1 billion to €2.4 billion. Timely inspection and targeted maintenance is therefore essential. Better underwater data play an increasing role in this. More accurate inspections enable better predictions about the life span of infrastructure and more targeted maintenance.

Until now, inspecting these objects has required the deployment of human divers. However, diving operations at locks and dams involve significant risks. For instance, divers can become trapped or entrapped underwater, as well as being swept along by strong currents. In the past, accidents, some of them fatal, have occurred during such work. Rijkswaterstaat therefore wants to limit diving activities as much as possible and aims for safer alternatives.

Practical trials in Farmsum

One such alternative is the deployment of underwater drones. Practical trials of these drones were conducted at the sea lock in Farmsum, Groningen, in late 2025. During a three-day trial, six companies tested systems in water with visibility of 30 to 50 centimetres. This visibility is representative of many Dutch inland waters.

This trial already showed that cameras are insufficient in such conditions. Many underwater drones are therefore equipped with acoustic sensors. These sensors make it possible to map both the shape and location of structures, detecting anomalies. Follow-up research should reveal whether the sensors can also be used to map exactly what these anomalies are, such as cracks. In addition, drones have been used to take thickness measurements of sheet piling and inspect mooring chairs, fenders and locks.

Analysing anomalies with 3D point cloud

TNO and Rijkswaterstaat point to the 3D point cloud created during the trials based on acoustic measurements. This point cloud makes it possible to analyse anomalies such as skewed sheet piling in detail.

The underwater drones were also found to be deployable to perform relatively complex work, which cannot be done by human divers. As an example, TNO and Rijkswaterstaat mention removing a car tyre stuck under a lock gate.

Replenishing and supporting human divers

Incidentally, the drones are not intended to replace human divers, but rather to complement and support them. TNO and Rijkswaterstaat, for instance, point out that divers have unique skills that drones cannot replicate. Think of being able to feel and listen. However, drones can support divers to a significant extent. For example, by exploring an area before human divers enter it, so that they enter the water optimally prepared.

In addition, drones can collect data that divers cannot gather. For instance, a drone can take measurements more systematically than humans and be equipped with equipment that a diver cannot carry.

Wouter Hoefnagel

Wouter Hoeffnagel is a freelance journalist and copywriter, with interests in both manufacturing industry, IT and the intersection between these topics. He writes a wide range of texts on these topics, ranging from background articles, interviews and news items to blog posts, white papers, case studies and website texts.