Riak’s scalable design makes it useful for data sets, like sensor data, that scale rapidly and are subject to heavy read/write loads. Many sensors collect and send data at a given interval. One way to model this in Riak is to use a bucket for each sensor device and use the time interval as a unique key (i.e. a date or combination of date and time), and then store update data as the value.
That data could then be queried on the basis of the interval. Alternatively, a timestamp could be attached to each object as a secondary index, which would allow you to perform queries on specific time interval ranges or to performMapReduce queries against the indexes.