British Royal College of Art counts on German digitization expertise
For the renowned Royal College of Art (RCA) in London, we have digitized, analyzed and assessed more than 1,600 real estate floor plans using the artificial intelligence that we developed for PORTER. The project is part of a study being carried out by the RCA on the housing market in the British capital, more precisely on the room standards and design characteristics of existing housing. PORTER floor plan digitization tremendously increased efficiency for the RCA: Those responsible for the project had the results from the 1,600 floor plan conversions in two weeks – 1.5 man-years had originally been projected for their completion!
- Floor plan digitization with the click of a button: simplest handling, fast and excellent results
- Efficiency: significant reduction in time – thanks to AI, roughly 97 percent less time was needed than projected
- Freed up resources and accelerated project implementation: Thanks to having the fastest possible digitization, analysis and assessment of floor plans as the foundation of the study, the Royal College of Art was able to continue their academic work earlier than anticipated.
- The RCA first needed to find a suitable service provider for floor plan digitization. The decisive reasons for working with XCYDE were not only the speed of floor plan recognition and the practicality of PORTER software, but also the anticipated quality of the results.
- The results needed to reliably deliver the data necessary for the study. PORTER was able to do just that: Our AI determined the room functions, dimensions and area of the real estate as well as connections between the rooms. Moreover, PORTER AI recognized elements such as windows, doors, bathroom fixtures and furniture.
- The aim was for the results to form the basis for developing alternatives, where applicable, to existing apartment descriptions and classifications (with regard to design and room attributes).
- With more than 1,600 floor plans, there was a strong likelihood that the quality of the floor plans might vary at least in some points. Where presentation in the floor plans was indeed insufficient, problems with presentation were to be expected, would need examination and would need to be manually resolved.
- Our AI had had only been trained to work with floor plans in the German-speaking realm and now needed to learn to recognize floor plans from the UK. It worked without any problems at all.
- The data output needed to be specifically adapted to suit the RCA: After all, the aim was to not only digitize the data, but to also academically analyze the results.
- The results from digitizing the 1,600 floor plans were available within three days. Reviewing the results and selective post-processing took us seven days. In total, only two weeks were necessary to produce excellent results.
- The AI is trained so well that it effortlessly recognized the 1,600, largely classic apartment floor plans. It was only in exceptional cases – when presentation on the floor plans was insufficient – that we needed to make manual improvements.
- It goes without saying that as with all customers, we remained in close and constant contact with the Royal College of Art and could be reached at all times to answer questions. As a result, those responsible for the study from the world’s best university for art and design were well and truly impressed by the speed of the software and the quality of the results.