Electric vehicle powertrain calibration

A method includes generating, for an electric vehicle powertrain, a calibration map providing a mapping from values of a set of context variables to mapped values of a set of decision variables. The context variables have values representing the vehicle's operational conditions, and the decision variables have values representing powertrain parameters adjustable by an ECU. The method includes determining the mapped values 572 in dependence on outputs of a Gaussian process model 570 arranged to predict values of a performance characteristic at points in an input space having dimensions corresponding to at least one context variable and at least one decision variable. A covariance function of the Gaussian process model depends on input space points in accordance with a specified physical relationship between the performance characteristic and at least a chosen variable of the decision or context variables. The performance characteristic may be a torque value, current advance angle and there may be more than 1 Gaussian process model.