.
The cement industry increasingly utilizes Alternative Fuels such as rice husk, fabric, paper waste, and polythene materials to reduce fossil fuel consumption and environmental impact. These fuels are mechanically preprocessed and transported via conveyor systems prior to combustion.However, due to the unregulated nature of waste derived fuels, the Alternative fuel mix can randomly contain oversized unwanted objects, including large metallic components, non-ferrous metals, and rigid contaminants. Although magnetic separators are commonly employed during pre processing, they are primarily effective at removing small ferrous metal particles, while large metallic objects, non-ferrous metals, and rigid contaminants frequently bypass separation stages. Such unwanted objects can cause severe blockages, conveyor damage, and costly production downtime. This paper presents an automated vision based detection and control system for identifying blockage objects on that alternative fuel. A YOLOv10 based deep learning model is deployed using real conveyor imagery collected from an operational cement plant. Upon detection, the system automatically stops the conveyor and issues alarms through a remote monitoring interface, enabling timely operator intervention. Experimental results demonstrate a detection accuracy of 95 %, confirming the effectiveness of the proposed system as an intelligent secondary safety layer that enhances operational reliability in cement plant alternative fuel handling.
This article exploits a 2018 reform of financial incentives for teachers to work in French disadvantaged schools. Based on administrative personnel records from an educational authority, it evaluates the impact of those incentives on stated preferences to move to such schools. This quasi-experimental source of variation of financial incentives, namely a gradual doubling of the annual stipend from €2,300 to €4,600 in these schools, enhanced their attractiveness. The estimated effect on teachers’ desired mobility is substantial and equivalent to a 34-minute reduction per trip. However, this effect is heterogeneous among teachers, notably higher for those with less experience and who already work in such schools.
The cement industry requires accurate and efficient quality control methods to ensure structural safety and material performance. This study proposes an integrated AI- and IoTbased framework for automated cement quality assessment. The system combines four main components: YOLOv9-based color particle identification for clinker microscopy analysis, IoTbased ultrasonic sensor measurement with U-Net-based crack detection for cement cube testing, multi-output cement strength prediction using an ensemble of XGBoost and LightGBM models, and automated clinker phase classification using MobileNetV2 with transfer learning. The system was validated using industrial datasets obtained from INSEE Cement (Private) Limited, Sri Lanka. Experimental results achieved 88% validation accuracy for particle identification, 90 % validation accuracy for crack detection, 88 % validation $\mathbf{R}^{\mathbf{2}}$ score for multi-age strength prediction, and 91 % validation accuracy for clinker phase classification. The proposed framework enables faster and more reliable cement quality assessment suitable for industrial deployment.
The development of peer-to-peer short-stay accommodation, mediated by online platforms, is a major phenomenon in the tourism and hospitality industry. This paper explores how it is traced in national accounts, especially for France, the largest European tourist market. The current methodology fails to fully embrace this activity and its associated production. A generic, more comprehensive approach to considering property owners as notional unincorporated enterprises is proposed. This is applied to France by exploiting a variety of sources. Compared to current methods, the output and productivity of the accommodation industry can accordingly be significantly revised upward. A unified approach can also enhance the comparability of this industry across countries.
