AI-Enabled Predictive Systems for Women Safety in Smart Cities

Abstract

Recent progress in Artificial Intelligence (AI) has made it possible to create smart systems that make smart cities safer for women. This paper offers an extensive evalua- tion of current AI-driven predictive frameworks, emphasizing their functionalities, constraints, and prospective advancements in proactive threat identification and mitigation. The review looks at a number of different methods, such as surveillance systems based on deep learning, gesture and voice recognition techniques, predictive crime mapping, mobile safety apps, and WiFi-based models for recognizing human activity. Research that combines technologies like YOLO, Res Net, Open Pose, BiL STM, and CNN-GRU shows that it is possible to find distress signals, suspicious behavior, and environmental risks in real time. The paper additionally discusses about how smart infrastructure solutions like intelligent street lighting, geospatial safety analytics platforms, and crowdsourced safety scoring systems can make cities safer for women. There are also talks about privacy- preserving machine learning and explainable AI frameworks to deal with ethical and transparency issues that come up with large-scale surveillance systems. The paper identifies important research gaps based on the literature that was reviewed. These gaps include the need for unified multimodal systems, zero- device safety mechanisms, and better integration with smart city infrastructure. The study concludes that AI-driven predictive systems, when integrated with ethical safeguards and urban planning strategies, can substantially improve women’s security, emergency responsiveness, and inclusivity in forthcoming smart cities.