UAV-CPSs as a test bed for new technologies and a primer to Industry 5.0
Ana Carolina Borges Monteiro, Reinaldo Padilha França, Vania Vieira Estrela, Sandro R. Fernandes, Abdeldjalil Khelassi, R. Jenice Aroma, Kumudha Raimond, Yuzo Iano, Ali Arshaghi · 2020
The extensive Cloud pool of resources and infrastructure can deliver significant improvements to Unmanned Aerial Vehicle (UAV) Cyber-Physical Systems (UAV-CPSs) relying on data or code from a network to operate, but not all sensors, actuators, computation modules and memory depots from a single fixed structure. This chapter is organised around the potential benefits of the Cloud: (1) Big Data (BD) access to visual libraries containing representations/descriptive data, images, video, maps and flight paths, (2) Cloud Computing (CC) functionalities for Grid Computing (GC) on demand for statistical analysis, Machine Learning (ML) algorithms, Computational Intelligence (CI) applications and flight planning, (3) Collective UAV Learning (CUL) where UAVs share their trajectories, control guidelines and mission outcomes and (4) human-machine collaboration through crowdsourcing for analysing high-dimensional high-resolution (HDHR) images/ video, classification of scenes/objects/entities, learning and error correction/ concealment. The Cloud can also expand UAV-CPSs by offering (a) data sets, models, all sorts of literature, HDHR benchmarks and software/hardware simulators, (b) open competitions for UAV-CPS designs with Open Source Hardware (OSH) and (c) Open-Source Software (OSS). This chapter talks about some open challenges and new trends in UAV-CPSs.