Mullite is a versatile material used in traditional and advanced ceramic, due to low thermal expansion, high thermal shock and creep resistance. The production of waste from ore industry is a problem that is increasing nowadays. The kaolin processing industry produces residue rich in Al2O3 and SiO2. These oxides are good precursors to produce mullite. The aim of this work is to use the residue from kaolin industry to produce mullite ceramic bodies. It was studied alumina, clay and residue from kaolin processing as raw materials. The material was dried and pressing in uniaxial press (30MPa). The samples were sintered at temperatures of 1450oC, 1500oC, 1550oC and 1600oC. The ceramic bodies were characterized by X-ray diffraction. The density was measured by Archimedes method and the flexural strength by the three point bending technique. The results showed that is possible to produce mullite with high amount of waste from kaolin processing.

Internet-of-Things (IoT) applications based on Artificial Intelligence, such as mobile object detection and recognition from images and videos, may greatly benefit from inferences made by state-of-the-art Deep Neural Network(DNN) models. However, adopting such models in IoT applications poses an important challenge since DNNs usually require lots of computational resources (i.e. memory, disk, CPU/GPU, and power), which may prevent them to run on resource-limited edge devices. On the other hand, moving the heavy computation to the Cloud may significantly increase running costs and latency of IoT applications. Among the possible strategies to tackle this challenge are: (i) DNN model partitioning between edge and cloud; and (ii) running simpler models in the edge and more complex ones in the cloud, with information exchange between models, when needed. Variations of strategy (i) also include: running the entire DNN on the edge device (sometimes not feasible) and running the entire DNN on the cloud. All these strategies involve trade-offs in terms of latency, communication, and financial costs. In this article we investigate such trade-offs in real-world scenarios. We conduct several experiments using deep learning models for image-based object detection and classification. Our setup includes a Raspberry PI 3 B+ and a cloud server equipped with a GPU. Different network bandwidths are also evaluated. Our results provide useful insights about the aforementioned trade-offs. The partitioning experiment showed that, overall, running the inferences entirely on the edge or entirely on the cloud server are the best options. The collaborative approach yielded a significant increase in accuracy without penalizing running costs too much.