The paper introduces the Cougar approach to in-network query processing in sensor networks, which aims to address the limitations of traditional centralized data processing systems. Existing sensor networks, which are often preprogrammed and send data to a central frontend for offline querying, have two major drawbacks: they cannot be dynamically reprogrammed and consume significant battery power. The Cougar approach uses declarative queries to generate efficient query plans for in-network processing, reducing resource usage and extending the network's lifetime. The paper outlines the architecture of a sensor data management system, including a query layer with a query proxy on each sensor node, and discusses open research problems such as aggregation, query languages, query optimization, catalog management, and multi-query optimization. The authors also review related work in ad-hoc and wireless routing, power-aware networking, distributed database systems, and adaptive query processing. The goal is to develop a robust and efficient system for managing sensor networks, leveraging the database community's expertise.The paper introduces the Cougar approach to in-network query processing in sensor networks, which aims to address the limitations of traditional centralized data processing systems. Existing sensor networks, which are often preprogrammed and send data to a central frontend for offline querying, have two major drawbacks: they cannot be dynamically reprogrammed and consume significant battery power. The Cougar approach uses declarative queries to generate efficient query plans for in-network processing, reducing resource usage and extending the network's lifetime. The paper outlines the architecture of a sensor data management system, including a query layer with a query proxy on each sensor node, and discusses open research problems such as aggregation, query languages, query optimization, catalog management, and multi-query optimization. The authors also review related work in ad-hoc and wireless routing, power-aware networking, distributed database systems, and adaptive query processing. The goal is to develop a robust and efficient system for managing sensor networks, leveraging the database community's expertise.