Performance Analysis of an Autonomous Cart Trolley for the Disabled: An ANFIS Based Evaluating Methodology

The development of mechanical systems, specifically autonomous systems, is popular in the literature of applied intelligence, robotics being one of them. This research is focused on the performance analysis of developed Sensor Controlled Cart Trolley. Research in this field has become necessary to make life easier for disabled persons and aged people in our society today. To the best of the authors' knowledge, no researcher has tested the intelligence of a developed cart-trolley using meta-heuristic techniques and hybrid creative algorithms. This is necessary to predict the intelligence and level of performance of an autonomous system. This research work is an exposition of the strength of a hybrid algorithm in terms of the performance analysis of a developed trolley. The hybrid algorithm implemented is the Adaptive Neuro-Fuzzy Inference System (ANFIS) model, using data obtained by testing a cart-trolley loaded with sand. The total intelligence test (TIT) of the autonomous cart-trolley was tested; the rough trial observed gave TIT of 45%. The first trial observed from the ANFIS rule viewer gave an average TIT of 52%. A second trial was performed, which gave an average of 59% TIT rating, and the final test gave a high TIT rating of 68.9%. The performance test is an indication that the developed sensor-controlled material handling device gave a satisfactory performance.