A perennial crisis of growing waste accumulations and natural resource depletion are major challenges for most parts of the world today, of which many have explored recycling as a solution. However, recycling is impeded by the need to sort waste by material types before processing. Additionally, the composition of municipal solid waste is so random such that the processes cost more than the value of recyclables. This research aimed at increasing the quantum of waste recycled, by using waste sorting machines at large institutions such as universities and hospitals. Unlike municipal solid waste, the composition of the waste from such institutions was less random and could be analysed through waste characterization and sorting. Waste characterization was conducted for a university and the data directly integrated into the design and subsequent calibration of a customized and automated waste sorting machine. The modular and calibratable machine capable of sorting 200kg of waste per hour was designed, with an estimated cost of USD 5,966.72. The automated design was sound and sustainable, based on the simulation and stress analysis carried out, capable of achieving 12m travel distance per item and at least 85% accuracy in sorting, with Von Misses stresses not exceeding 0.168 MPa.