This paper is focused on a machine-learning approach for estimating human energy expenditure during sport and normal daily activities. The paper presents technical feasibility assessment that analyses requirements and applicability of smart phone sensors to human energy expenditure. The paper compares and evaluates three different sensor configuration sets: (i) a heart rate monitor and two standard inertial sensors attached to the users thigh and chest; (ii) a heart rate monitor with an embedded inertial sensor and a smart phone carried in the pocket; and (iii) only a smart phone carried in the pocket. The accuracy of the models is validated against indirect calorimetry using the Cosmed system and compared to a commercial device for energy expenditure SenseWear armband. The results show that models trained using relevant features can perform comparable or even better than available commercial device.