Mass air flow

Modern vehicles use a mass air flow (MAF) sensor for optimum fuel delivery. The mass air flow is used to determine the mass of air entering the engine. The air mass information is necessary to
calculate and deliver the correct fuel mass to the engine. Air is a gas, and its density changes as it expands and contracts with temperature and pressure. Consequently, in automotive applications the air density continuously varies with the vehicle's operating environment.

When a MAF is used in conjunction with an exhaust gas oxygen sensor,
the engine's air/fuel ratio can be controlled very accurately. The MAF sensor provides the open loop predicted air flow information (the measured air flow) to the engine's ECU, and the EGO sensor provides
closed loop feedback in order to make minor corrections to the predicted air mass. Common drawbacks of the widely used hot wire MAF sensor include high cost and sensor degradation resulting from oil
and dirt build-up degrading sensor accuracy.

When a MAF sensor gets contaminated, it skews the transfer function
such that the MAF overestimates airflow at idle (causes the fuel system to go rich) and underestimates airflow at high airflow (causes the fuel system to go lean).
Neural network virtual sensors have been shown to accurately reproduce MAF sensor functionality at a fraction of the cost with none of the sensor degradation issues. Mosaix provides high performance neural network based MAF virtual sensors.