As digital technologies become prevalent and embedded in the environment, "smart" everyday objects like smart phone and smart homes have become part and parcel of the human enterprise. The ubiquity of smart objects that produce ever-growing streams of data presents both challenges and opportunities. In this paper, we argue that extending these data streams, referred to as "predictive analytics", provides a solid basis for the design and development of IS artefacts that can generate additional value. Subsequently, we introduce a model for Designing Information Systems with Predictive Analytics (DISPA), extending Design Science Research specifically towards predictive analytics. The model is evaluated based on a case study of MAN Diesel and Turbo, a leading designer of marine diesel engines. The case illustrates that the framework provides useful guidelines for developing environment-specific sensor based predictive models that can out-perform the traditional state of the art predictive methods especially in volatile and uncertain environments.