Roads have multiple effects on wildlife, from animal mortality, habitat and population fragmentation, to modification of animal reproductive behaviour. Monitoring roadkill is expensive and time-consuming, and depend mainly on volunteers. Thus, cheap, easy to implement, and automatic methods for detecting roadkill over larger areas and over time are necessary. We present results from the research project Life LINES, where we developed a cheap and efficient system for detecting amphibians and small birds roadkill using computer vision techniques. We present here the Mobile Mapping System 2, an improved version of the Mobile Mapping System 1 developed during the Roadkill-project and presented in previous IENE congresses. We have successfully reduced the size and energetic consumption of the MMS, so now the device can be attached directly to the back of any car. The MMS2 is composed by several cameras (multi-spectral, visual with 3D laser technology, and high definition). The algorithms were trained with previous collected pictures of road-killed amphibians and small birds. We have tested all images using the Haar Cascade algorithm from the OpenCV library, which provided high rate classification results. We tested the MMS2 in three conditions: a control test with plastic models of amphibians and birds in a small road; a control test with collection specimens of amphibians and birds; and a real test on a 30 km road survey in Southern Portugal. The MMS2 has been developed using low cost components with the idea of saving funds, time and personal resources for wildlife preservation.