Abstract: A camera in a facility acquires an image of a user performing an action, such as shopping. These images may include potentially identifying information, such as an image that includes the user's face. To safeguard privacy of the user, a modified image is generated by changing one or more characteristics of the acquired image. The modification affects the entire image, rendering the resulting image unsuitable for another person to identify the user. For example, a contrast value for the image may be changed, rendering the user unrecognizable. This modification removes the potentially identifying information before later processing takes place. For example, the modified image in which the user's face is obscured may retain other non-identifying information used to train an artificial neural network to recognize actions such as pick or place of items in the facility.

Abstract: The present disclosure is directed toward a storage compartment vehicle (“SCV”) for use in the transport of items for delivery to one or more customers and/or retrieval of items from one or more customers. The SCV may include a plurality of storage compartments that are individually accessible, each of which can contain one or more items. The SCV may navigate to a delivery location and customers may retrieve ordered items from the storage compartments of the SCV. Likewise, customers may return items into storage compartments of the SCV.

Abstract: The present disclosure is directed toward a storage compartment vehicle (“SCV”) for use in the transport of items for delivery to one or more customers and/or retrieval of items from one or more customers. The SCV may include a plurality of storage compartments that are individually accessible, each of which can contain one or more items. The SCV may navigate to a delivery location and customers may retrieve ordered items from the storage compartments of the SCV. Likewise, customers may return items into storage compartments of the SCV.