Systems and techniques for assessing a customer premises equipment device

An electronic device, comprising:

• one or more processors;
• and
• memory storing one or more programs configured to be executed by the one or more processors, the one or more programs including instructions for:
• determining, based upon information comprising functional parameters and diagnostic data for a first customer premises equipment (CPE) device, a health score for a second CPE device, wherein the first and second CPE devices are both connected to a network corresponding to a particular service provider and the health score is determined using a machine learning system that bases the determination of the health score at least in part on the functional parameters and diagnostics data of the first CPE device compared to the functional parameters and diagnostics data of the second CPE device;
• based upon a comparison between the health score and a threshold, determining whether the health score indicates that the second CPE device is defective;
• in response to a determination that the health score indicates that the second CPE device is defective, outputting second information indicating that the second CPE device is defective and needs to be replaced and returned;
• and
• in response to receiving an indication that the second CPE device was tested and determined not to be defective, updating the machine learning system used to determine the health score with a training example including diagnostic data of the second CPE device and that the second CPE device is not defective.