Cloud based method for improving residual value estimation for end-of-life product recovery

Manufacturing companies that take back products at the End-Of-Life (EOL) face a decision whether to recycle, remanufacture, repair, refurbish or scrap that product. One of the first steps is to estimate the product's residual value, primarily on the basis of age and a cursory visual inspection. However, there is a high degree of variability in actual value, even for products of the same age. This is due to the variability of environmental and use conditions to which the product has been exposed by the consumer. This paper proposes a predictive model that uses data obtained from sensors which is stored on the cloud, throughout a product's lifecycle. This data is used to more accurately estimate the value at the EOL. The model recommends the EOL solution for an individual product based on the quality level and demand for refurbished products. An illustrative cell phone example is presented, which tracks the condition of each phone during its lifecycle. Simulation is performed to obtain the residual value distribution based on predictive indicators from sensors including accelerometer data to monitor number of free falls and impact, number of battery lifecycles, humidity and temperature sensors. Results indicate improved residual value estimation due to minimizing the loss due to variance.