Common Reed (Phragmites Australis) is an abundantly available, sustainable construction material that can be found throughout the world. Reed is most commonly used for thatch roofing in Europe, providing insulation and a weather-tight surface. Elsewhere, traditional techniques of weaving and bundling reeds have long been used to create entire buildings. This research develops a new alternative to these techniques with the aim of showcasing the ability of reed to perform as structure, insulation, and cladding all at once. In the Netherlands, the availability of inexpensive imported reed has led to a decline in demand for Dutch reed for thatching. This is problematic as the management of reed beds is essential for nature conservation. The research aims to promote the use of Dutch reed for construction through utilising a digital production chain in order to reduce labour costs. Through an iterative process of designing from the micro to the macro scale and by experimenting with robotic assembly, the result is a reed-based system in the form of discrete components that can be configured to create a variety of structures. The project questions permanence in architecture, through the proposal of a series of nature observation structures for the National Park Duinen van Texel. The structures are intended to last ten to twenty years and utilise completely biodegradable materials which can be disassembled whenever necessary. This semi-permanence allows for future flexibility and ensures preservation of the natural environment. Physical prototyping and testing of the proposed robotic assembly process validates the approach.