Wavelength-shifting (WLS) materials contain molecules that absorb light and reemit at longer wavelengths. They can be used for light detection because they provide a large effective area for low cost and they are able to efficiently trap and guide light because of total internal reflection processes. We are currently developing such a WLS detector, considering two main designs: A single-shift design with one wavelength shift (tile) and a double-shift design with two wavelength shifts (tile and fiber). As photodetectors we use small Silicon photomultipliers (SiPMs) with a high photon detection efficiency (PDE) and single-photon sensitivity. The double-shift layout goes at the expense of detection efficiency. In this design however, light is channeled to the two ends of a fiber, thus requiring a reduced photosensitive area compared to the single-shift layout. We will present the results of our measurements and show that light traps and SiPMs together represent a promising alternative to PMTs in case of a non-focused light beam. For the special case of SWGO, the application of light traps is also motivated by a possible improvement of the gamma/hadron separation, using a one-chamber tank with an array of wavelength-shifting light traps instead of a (two-chamber) tank with PMTs. Besides SWGO, new WLS detectors could also constitute useful and cheap technology for other experiments and use cases. The contribution summarizes our motivation and efforts to build a light trap detection module and to characterize its properties in terms of costs, temporal performance and detection efficiency.