Abstract:
Vehicle based logistics are hinged on their ability to timely deliver goods, services, and people. The classical expression of “time is money” comes alive in the logistics industry yielding potentially huge financial and health consequences in case of missing deadlines. This is especially the case for time sensitive pharmaceuticals, delivery of perishable goods, delivery of people travelling, delivery of services in fault fixing/recovery sector. All these use cases motivate the need for an immutable, secure, and immortalized process of tracking time. To solve this challenge, this thesis presents prototype-based research that integrates the 4th industrial revolution technologies of vision Internet of Things (IoT), Artificial Intelligence (AI) Object detection, Optical Character Recognition (OCR) and blockchain. The developed prototype features a Raspberry-PI board embedding a camera, an Artificial Intelligence (AI) model to recognize plate letters from the image and a crypto wallet to sign the logging of plate number and time events on the NEAR blockchain, an emerging sharded, layer-one blockchain that uses proof-of-stake and is simple to use, secure and scalable. The effective operation of the developed prototype has been validated inside a campus parking and shows an accuracy range of 92-98%. The benefits of transparency, security, and immutability of the blockchain combined with the intelligence, data capture, and processing of IoT will help to improve the development of accountable solutions trusted by all different logistic stakeholders. As one of the main achievements of this research, a related paper has just been accepted for publication in the conference 8th International Conference on Machine Learning Technologies (ICMLT 2023).
