Product 'in-use' context identification using feature learning methods

Usage context is considered a critical driving factor for customers' product choices. In addition, the physical use of a product (i.e., userproduct interaction) dictates a number of customer perceptions (e.g. level of comfort, ease-of-use or users' physical fatigue). In the emerging Internet-of-Things (IoT), this work hypothesizes that it is possible to understand product usage while it is 'in-use' by capturing the userproduct interaction data. Mining the data and understanding the comfort of the user adds a new dimension to the product design field. There has been tremendous progress in the field of data analytics, but the application in product design is still nascent. In this work, application of 'feature learning' methods for the identification of product usage context is demonstrated, where usage context is limited to the activity of the user. Two feature learning methods are applied for a walking activity classification using smartphone accelerometer data. Results are compared with feature-based machine learning algorithms (neural networks and support vector machines), and demonstrate the benefits of using the 'feature learning' methods over the feature based machine-learning algorithms.