Classic urban design theories and research methods have been limited to qualitative approaches, such as subjective intuition and small-scale surveys(Batty, 2013; Kvan, 2020). With the emergence of new urban science, adopting big data, virtual reality, and wearable sensors, it is possible to achieve a precise analysis of people’s perception and behavior of urban space(Offenhuber and Ratti, 2014). In this way, not only will new insights in urban study be inspired, but also high-quality, human-scale space design can be generated(Shen et al., 2017; Ye et al., 2019). A systematic review was conducted to map the emerging literature in related fields(Townsend, 2015). With the keywords “urban design”, “big data” and “new urban science,” more than 200 representative journal articles published on Web of Science in recent years were selected as the dataset. There are three new research directions emerging on the leading edge of the literature (Fig.1). In this context, a systemic brand-new framework for computational urban design can be proposed, integrating big data, machine learning algorithms, and geographic design to achieve a refined analysis from a human-oriented perspective. The new algorithm is applied to the depth of the whole process of urban design, with an emphasis on fusing science, technology, and design depth, in the form of quantitative calculation to support a better design(Miao et al., 2018; Wilson et al., 2019; Tunçer, 2020). In order to deal with complex urban design projects, computational urban design has to involve three major directions: data-informed, evidence-based, and algorithm-driven (Fig.2). This paper took the Shanghai Urban Design Challenge as an example of computational urban design in practice(Fig.3). First, the data-informed urban design is helpful to identify problems and formulate strategies. Multi-source city data, such as POIs data, and AutoNavi map path planning API, helps refine the needs, functions, and characteristics of the citizens(Fig.4). The evidence-based urban design with specific inquiries helps to select the design intervention point and control spatial elements. Virtual reality technology, wearable biosensors and the visualization SP method are combined to measure space utility. The final algorithm-driven part can use sDNA, UNA and virtalvizer to simulate spatial elements, to perform morphological analysis and vitality evaluation, and to further adjust the design(Fig.5). This paper introduces systematic and scientific thinking into urban design practice and controls the whole process and all directions of the design process. It can be seen that the application of this framework can make the entire design process more automatic, efficient, and robust under the existing constraints. New design tools enable urban research to directly support the most central part of urban design: scheme generation. More importantly, through simulation prediction, the effectiveness of the scheme is evaluated, and the designer and decision-maker can obtain real-time feedback and check the scheme accordingly. In addition, it helps the perspective of urban design expanding from the previous “top-down” perspective to the “bottom-up” perspective, transforming from a two-dimensional to a three-dimensional, obtaining real-time feedback and optimizing design accordingly, emphasizing embodied perception and hu-man-oriented scale.