Large-scale sensor data distributions and knowledge inferences are major challenges for cognitive-based distributed storage environments. Cognitive storage sinks play an essential role in addressing these challenges. In a data-concentrated distributed cognitive sensor environment, cognitive storage sinks regulate the data distribution operations and infer knowledge from the large amounts of sensor data that are distributed across the conventional sensors. Embedding cognitive functions in conventional sensors is unreasonable, and the knowledge-processing limitations of conventional sensors create a serious problem. To overcome this problem, we discuss a cognitive co-sensor platform across a large-scale distributed environment in which the cognitive storage sinks are the regulating authorities of data distributions and knowledge inferences and the conventional sensors are the data sources. In our work, we also discuss a distributed data distribution system for effective data distribution across the distributed cognitive storage sinks and a distributed knowledge inference system that infers useful patterns for building knowledge intelligence.