You are designing a schema for a Cloud Bigtable instance that will store sensor data for a smart home application. Each sensor generates readings every second, and you plan to store the data using the sensor ID and timestamp as the row key. During initial testing, you notice that reads and writes are bottlenecked because certain sensors are generating significantly more data than others, causing uneven load distribution across the cluster. You need to redesign the row key to improve performance and load balancing while keeping the row keys human-readable. What should you do?