Architects and urban planners still rely on physical models in the design process, particularly for urban contexts. However, developing such models can be time-consuming and resource-intensive, despite the abundance of digital information available. This study investigates the feasibility of using pin-type models, specifically in the concept and design phase of mass modelling and analysis for urban, rural, and complex building sites. The research uses a combination of literature review, design-based research, and interviews, leading to the discovery of a promising solution. Existing pin-type models in the market or universities either come at a high cost or lack the necessary resolution to generate detailed urban models. However, this study identifies a method that significantly reduces costs while achieving higher pin resolution, although with some compromises in generation time and holding power compared to current models. The innovative rubber layer, also known as the state layer, plays a crucial role in the machine's advancements. This layer enables the machine to keep pins in place at a lower cost but with lower holding power. Additionally, leveraging technology from the 3D printing and CNC manufacturing domains, the machine incorporates multiple individually moving motors that can rapidly set up to 40 pins in one motion, reaching speeds of up to 1200 pins per minute. Making it possible to generate large pin sets in less then an hour, meaning it would be faster than traditional methods like 3D printing or manual foam cutting. Furthermore, the research emphasizes the central role of code in this study. Integrating all the necessary steps into a single code can generate an urban area within minutes. The program facilitates a user-friendly interface for selecting locations, automatically downloading height maps from the AHN (Actueel Hoogtebestand Nederland), converting the data into a digital model compatible with the machine, and exporting it as a G-code file. The machine and code also support displaying additional digital information, such as 3D models and images. Combining the improved machine and optimized code demonstrates the potential to create physical urban context models from digital data at a relatively low cost compared to other devices. With working times of minutes instead of hours or even days, compared to traditional model-making methods. Although this research shows considerable potential, there are several challenges that demand attention and additional investigation. The ultimate determination of whether the final product will require a $10,000 or $20,000 investment hinges upon the availability and cost of visible pins. Moreover, the current conversion of rotation to Bowden cables is suboptimal, and further enhancements are necessary. Additionally, the code could benefit from optimization to ensure compatibility with slower computers, thereby improving its speed and efficiency. Furthermore, the code should be extended to enhance its compatibility with 3D models, and ideally, with BIM models as well. It is essential that the code incorporates a function enabling the combination and selection of data, facilitating the movement of only the necessary sections of the machine. With this research, a solid foundation has been established for further expansion and exploration in this field. The successful development of a working machine code paves the way for practical testing and implementation. Although there are some mentioned additions that could enhance the research, the limitations of time prevented their full exploration in this study. Nonetheless, this research opens up exciting possibilities for future advancements and applications in the field.