Sensor Stream Moving Average

You are given a stream of events representing sensor readings. Each event is a record with the following fields:

• sensor_id: a unique identifier for the sensor (string or integer).
• timestamp: the time at which the reading was taken (you can assume a Unix timestamp in seconds).
• value: a numeric value recorded by the sensor.

Your task is to design and implement a system that processes this stream of events in real time and calculates a moving average of the sensor values over a fixed time window (e.g., the last 60 seconds) for each sensor. Note that:

1. Events can arrive out-of-order (i.e., a later event might have an earlier timestamp than one that has already been processed), but you should still calculate correct moving averages.
2. Your solution should be efficient in terms of time and space, as the stream can be large and there may be many sensors.
3. Include functionality to query the current moving average for any given sensor.
4. Assume that sensor events can continue to come indefinitely; design your system to handle this gracefully without exhausting system resources.

Implement your solution in a programming language of your choice. Ensure that your code is clean, modular, and well-documented.