Manikandan Palanichamy

In 1874, the French science fiction writer Jules Verne imagined a world where water would serve as fuel. This vision, more than 150 years later, is now closer to reality than fiction. Green hydrogen, produced by splitting water using renewable electricity, has emerged as one of the most promising paths for decarbonizing hard-to-abate sectors such as steel, chemicals and heavy transport. Global electrolytic hydrogen capacity reached 2 GW in 2024, with low-emissions hydrogen production projected to grow fivefold by 2030. However, the economics remain challenging. Green hydrogen still costs 1.5–7 times more than fossil-derived alternatives, and the vast majority of global hydrogen production today still depends on natural gas or coal. Artificial intelligence (AI) is increasingly recognized as a critical enabler for bridging this gap. AI can support the optimization of production, enabling predictive maintenance and enhancing safety across the hydrogen value chain. However, a fundamental question remains underexplored: Can we trust AI to make decisions in hydrogen environments where errors are not just costly, but potentially catastrophic? This editorial examines the opportunities AI presents for hydrogen energy systems, the safety challenges it must overcome and the regulatory frameworks urgently needed to govern its deployment.

Reliable degradation prediction for Proton Exchange Membrane Fuel Cells (PEMFCs) is essential for durability assessment and condition-based maintenance. This study proposes a physics-informed hybrid Transformer–LSTM framework for PEMFC Health Index (HI) prognostics under WLTP and NEDC operating profiles. A voltage-dependent HI was developed from experimental degradation data, and degradation behaviour was decomposed into reversible and irreversible components using Savitzky–Golay smoothing. Short-term validation showed that the Transformer underperformed the LSTM under both operating conditions. Long-term Prediction Horizon analysis demonstrated reliable prediction horizons of ~400 h under WLTP and ~500 h under NEDC, with low uncertainty growth rates and near-ideal uncertainty calibration. Activation and humidification enabled partial performance recovery, increasing open-circuit voltage from 10.5 V to 11.4 V and recovered power output to ~876 W.

Digital transformation in agriculture prioritized efficiency over climate adaptation, and such systems struggle when conditions shift beyond historical patterns. Current research remains fragmented while perceiving AI as pure automation and overlooking human cognition, collective networks, and ethics. This chapter introduces the Quad Intelligence Framework: Computational Intelligence (AI optimization), Human Cognitive Intelligence (adaptive judgment), Collective Intelligence (farmer networks), and Ethical Intelligence (accountability). Ten global implementations were analyzed across monitoring, early warning, and crop management, finding they favor efficiency over adaptation. Computational intelligence alone delivers short-term productivity but cannot adapt to climate uncertainty. Strategic integration of all four dimensions transforms efficiency-based systems into climate-resilient instruments by incorporating farmer's climate knowledge, collective networks, and sustainability thinking. The chapter provides roadmaps for transitioning to resilience-oriented digital agriculture.

LoRaWAN (Long Range Wide Area Network) contributes as a key technology in Industrial Internet of Things (IIoT). LoRaWAN receives consistent industrial attention due to its far-reaching long-range communication capability, low power consumption, and easy-to-adapt architecture, which requires minimal infrastructure costs. This paper provides more insightful information about LoRaWAN in IIoT, illustrating how LoRaWAN optimises operational efficiency, safety, remote data collection, and sustainability through its dynamic features in industrial settings. Additionally, this paper explores the integration of LoRaWAN with emerging technologies, including artificial intelligence, machine learning, edge computing, and blockchain. While this paper emphasizes the potential of promising trends in the IIoT ecosystem, it also explores relevant challenges like scalability, security, and interoperability that demand the need for continued research. The comparative analysis of recent research works highlights key technical parameters. It is helpful to critically examine the trade-offs between performance, scalability, and security, and address them with novel solutions. In summary, this paper offers valuable insights for industry professionals and researchers seeking to leverage LoRaWAN to advance the Industrial Internet of Things (IIoT).

We survey delay testing of digital circuits. The advanced semiconductor process technology results in high-speed circuits in which the presences of delay faults are highly possible. Delay faults affect the timing behavior of the circuit when operated at its designated speed of operation. To detect such delay defects and guarantee the timing correctness of the circuit delay testing plays an important role. In this paper, we review several delay testing techniques with emphasis of path delay fault testing. Many innovations have been introduced since 1990s, both in delay fault models and test generation techniques. A review of this paper highlighting important delay test methodologies of the recent past with the assumption of its greater importance in the future than it has now. We also applied the path delay fault testing method in hardware security. Experimental results on several benchmark circuits show that we can achieve 12.7%-19.6% variation in statistical path delay fault coverage in the infected integrated circuits.