General Motors (India)

Understanding spatiotemporal changes in extreme rainfall is critically important in monsoon-dominated regions. This study investigates trends in heavy (HR), very heavy (VHR), and extremely heavy (EHR) rainfall events across Middle Gujarat and Saurashtra, India, using quality-controlled daily ERA5 reanalysis data for the period 1980–2022. Six climatically diverse districts Ahmedabad, Dahod, Panchmahal, Amreli, Junagadh, and Rajkot — were selected for analysis. Rainfall trends were assessed using linear regression and Sen's slope estimator, with statistical significance evaluated through the non-parametric Mann–Kendall test. Results reveal spatially heterogeneous rainfall behavior across the study region, with weak to moderate increasing tendencies observed in HR events and comparatively clearer increasing trends detected in VHR events at selected stations. EHR events, though rare, exhibit notable episodic increases in recent decades, particularly over urban and semi-arid areas. However, the majority of detected trends do not attain statistical significance at the 95% confidence level, largely attributable to strong interannual variability and the sporadic nature of extreme rainfall occurrences. Despite limited statistical significance, the observed shift toward higher-intensity rainfall events highlights growing hydroclimatic vulnerability across Gujarat, underscoring the urgent need for impact-based risk assessment, improved urban drainage design, and climate-resilient water resource planning strategies.

The Probation of Offenders Act, 1958 (POA) was the first Act which started to shift the Indian penal system from punitive to rehabilitative. This Act had an objective to combat prison overcrowding, staving off the criminal socialization process, and giving offenders the chance at social reintegration. However, the Probation remains underused and the system viciously bypasses all the implementers. This paper approaches the study of probation from a doctrinal constitutional and international perspective, the case for which is built on comparative jurisprudence, institutional analysis, and comparative analysis. The paper posits that the system of probation in the country will necessitate synergistic reform across the spectrum of judicial education, standardized PSRs, the professionalization of probation officers, and the use of tech in judicial processes. This paper seeks to argue that probation, when viewed socially and philosophically through the lens of the constitution, is a constitutional right to not be jailed, and should be coupled with international legislation, the Tokyo Rules, which should call for reformative probation instead of punitive incarceration.

Automatic Speech Recognition (ASR) systems have undergone tremendous advancement over the last few years but remain challenged with errors in transcription and readability, particularly in spontaneous speech. In this paper, a hybrid ASR system that employs the T5 transformer model in postediting is introduced for improving transcription quality and coherence. Our methodology focuses on utilizing a conventional ASR pipeline to obtain preliminary transcripts followed by a fine-tuned version of the T5 model, which refines errors and makes the transcript readable. We experiment with the system on several corpora, producing significant gains in word error rate (WER) and linguistic clarity. Experimental metrics indicate that the hybrid method not only surpasses baseline ASR models in precision but also achieves greater readability, rendering it especially appropriate for realistic applications like subtitles, meeting records, and virtual assistants. Further, we also investigate the versatility of the model to accommodate variability in speaking behavior and domain terms. The system we suggest points towards the capabilities of transformer-based post-processing in transforming ASR output to make it more human-like and contextually correct. Our research also discusses the computational efficiency of the process, which ensures its practicality for real-time deployment. We also test the robustness of the model over various conditions of noise and speech accent and prove its generalizability.

Abstract Securing and transporting cargo is common in vehicles of all types; however, the current cargo retention approaches (e.g., cargo nets, storage bins, and elastic cords) do not always provide adequate constraint to keep items securely in place while driving. One promising approach is to employ internally tiled pneumatic surface technology to design a moldable active cargo blanket, which can be draped over target objects, shaped into myriad forms, and rigidized on demand to ensure that the cargo items are effectively constrained. This article introduces a model-based approach for systematically designing pneumatically activated moldable active cargo blankets, providing tailorable moldability performance to cater to different vehicle segments, styles, and/or intended customer experiences. The architecture of moldable active cargo blankets comprises layers of low-profile and uniformly distributed rigid tiles within an airtight bladder, enabling transition from soft to rigid states as a function of vacuum pressure applied, providing moldability technology capability, which can be decomposed into three key technology subcapabilities: drapability, shapability, and rigidizability. The performance of each subcapability can be quantified in its respective operation states, draping, shaping, and rigidizing, by developing multiple engineering performance metrics characterizing each state. A half-factorial design-of-experiment investigating the relationship between the tile array design variables on these quantifiable metrics is conducted. A predictive modeling approach using empirical data to understand the mechanically complex behavior of moldable active cargo blanket is developed, relating tile array design variables to performance outcomes. This enables an algebraic tailoring method for selecting a specific set of tile array design variable values to balance the tradeoffs among metrics to obtain intended design outcomes, which is demonstrated through three distinct design contexts. The work in this article provides the enabling basis for the moldable active cargo blanket application as well as a more general technology basis on moldability.

Maintenance, production, and quality are the three main pillars of any industry. Maintenance is integral in helping the industry achieve its production targets without compromising product quality. The maintenance in the industry has evolved with highs and lows. Breakdown maintenance was one of the most common types of maintenance in the first industrial revolution. The breakdown maintenance later evolved into periodic maintenance and preventive maintenance in the second and third industrial revolutions. The advent of the fourth industrial revolution in 2010 enabled industries to venture into automation technologies like digital twins, artificial intelligence (AI), virtual reality, and many other allied domains. Predictive maintenance was one domain that rose to prominence after the introduction of the fourth industrial revolution. Predictive maintenance enables the industrial stakeholders to get a prior intimation of the failure of a machine or machine component using sensors, cyber-physical systems, AI, and other interdisciplinary domains. Senseye is a Siemens Company. They developed the Senseye PdM software that works on machine learning and AI to reduce the unplanned downtime of the machines. It uses the machine, maintenance, and maintainer behavior data to analyze the future health of the machinery and its critical parts that need attention. They published a report titled “The True Cost of Downtime” in 2022. The reports stated that large industrial establishments all across the globe had lost 3.3 million hours annually due to machine failures. These machine failures are almost equivalent to USD 1 Trillion. According to a report titled “Predictive maintenance and the smart factory,” manufacturers all across the globe face an annual loss of USD 50 Billion due to unplanned equipment downtime and costs involved in handling those issues. To reduce the losses of unplanned equipment downtime there is a strong need to implement predictive maintenance systems to supervise the health condition of the equipment. Over the last 5 years, hyperautomation has taken center stage in the manufacturing industry. Hyperautomation is a disciplined approach to automating industrial processes using novel technologies like data science, AI, digital twins, and many other technologies. This chapter provides an overview of the research papers published in the domain of the application of hyperautomation in predictive maintenance from 2015 to 2022.