Institute of Rural Management Anand

Corporate social responsibility (CSR) has assumed a crucial role in the modern business landscape, thereby moving beyond the realm of CSR as a voluntary engagement by firms. Businesses are expected to act responsively and be sensitive to the problems and issues faced by society at large. The development of Sustainable Development Goals (SDGs) by the United Nations in 2015 was a watershed event; the fulfillment of these goals requires concerted efforts from different stakeholders in society. There is enough evidence to believe that firms can, through CSR and otherwise, play a critical role in achieving these goals. Therefore, considerable research has taken place at the intersection of these two, providing valuable insights to researchers and practitioners. In connection with this line of research, the present study is a thematic review, the scope of which is to analyze the literature published on the role of CSR in achieving SDGs. The timeline for this search was between 2015 and 2024, with inclusion criteria of selecting articles focusing on CSR aimed at SDGs while excluding the standalone articles on CSR and SDGs. Literature highlights the importance of policy initiatives for strengthening the link between SDGs and CSR. The role of governments across the world in improving the nexus between CSR and SDGs to address the grand challenges faced by society is of crucial importance. Further, based on the extant literature, the study developed a framework to appraise the current body of knowledge and as an additional framework for future research. Substantial research highlights the role of CSR in achieving sustainability goals. CSR strategies work in line with the SDGs to generate goodwill from stakeholders and contribute to the overall development of society. The study suggests adopting a comprehensive approach integrating CSR strategies and sustainability models of business to improve both financial performance and social impact. Moreover, it is crucial for policymakers to strengthen the partnerships between businesses, higher education institutes, and the community to enhance the effectiveness of CSR initiatives.

Training deep neural networks and Physics-Informed Neural Networks (PINNs) often leads to ill-conditioned and stiff optimization problems. A key structural feature of these models is that they are linear in the output-layer parameters and nonlinear in the hiddenlayer parameters, yielding a separable nonlinear least-squares formulation. In this work, we study the classical variable projection (VarPro) method for such problems in the context of deep neural networks. We provide a geometric formulation on the Grassmannian and analyze the structure of critical points and convergence properties of the reduced problem. When the feature map is parametrized by a neural network, we show that these properties persist except in rank-deficient regimes, which we address via a regularized Grassmannian framework. Numerical experiments for regression and PINNs, including an efficient solver for the heat equation, illustrate the practical effectiveness of the approach.

In this paper, we introduce a new asymmetric weak metric on the Teichm & uuml;ller space of a closed orientable surface with (possibly empty) punctures. This new metric, which we call the Teichm & uuml;ller-Randers metric, is an asymmetric deformation of the Teichm & uuml;ller metric and is obtained by adding to the infinitesimal form of the Teichm & uuml;ller metric a differential 1-form. We study basic properties of the Teichm & uuml;ller-Randers metric. In the case when the 1-form is exact, any Teichm & uuml;ller geodesic between two points is also a unique Teichm & uuml;ller-Randers geodesic between them. A particularly interesting case is when the differential 1form is the differential of the logarithm of the extremal length function associated with a measured foliation. We show that in this case the Teichm & uuml;ller-Randers metric is incomplete in any Teichm & uuml;ller disc, and we give a characterisation of geodesic rays with bounded length in this disc in terms of their directing measured foliations.

Purpose This study compiles existing multidisciplinary research on additive manufacturing (AM) and circular economy (CE) to clarify how organisations can use AM to build capabilities that support CE objectives. It explains how AM adoption can enable these CE-oriented capabilities and summarises the main approaches firms can take to translate AM's potential into circular outcomes. Design/methodology/approach This research uses the 10R CE principles and Regenerate, Share, Optimise, Loop, Virtualise, and Exchange (ReSOLVE) framework to provide an AM-CE capabilities framework, based on which the authors provide propositions for future research. The primary themes in existing research, as well as the advantages and challenges of AM within the CE context, are discussed. Based on the findings, an AM-CE capabilities framework has been presented that helps understand the benefits of using AM and the obstacles that need to be overcome to utilize the technology for developing CE-based capabilities. The research uses the TCCM (Theory, Context, Characteristics, and Methodology) framework to pinpoint gaps in existing research on AM and suggest possible future avenues. Findings The review shows that AM can support CE goals, but its impact depends on the capabilities that the firms build around it. The study specifies three capabilities: (1) aligning product design and customer involvement with CE goals to reduce waste; (2) strengthening internal readiness through research and development (R&D), training and top management commitment; and (3) enabling adoption through policy support and stakeholder knowledge sharing across the value chain. These insights are consolidated in the AM-CE capabilities framework and translated into propositions for future research. Practical implications For researchers, the framework and proposition shift the conversation from “AM applications” to when and how AM helps build CE-based capabilities. For practitioners, the study highlights how CE outcomes from AM can be realized when firms connect customer needs and design choices with their CE goals, invest in R&D and workforce alignment, and actively collaborate with partners while navigating regulatory requirements. For policymakers, the study explains how clear standards and supportive interventions can reduce adoption friction and encourage more effective implementation. Originality/value The findings will assist practitioners in understanding AM's role in establishing CE-based capabilities for sustainable development. This research may help practitioners lower their carbon footprint by showing how to meet CE targets using a single technology (AM). By identifying hurdles and action items, decision-makers may better prepare for AM's advantages.

We present a geometric framework for Reinforcement Learning (RL) that views policies as maps into the Wasserstein space of action probabilities. First, we define a Riemannian structure induced by stationary distributions, proving its existence in a general context. We then define the tangent space of policies and characterize the geodesics, specifically addressing the measurability of vector fields mapped from the state space to the tangent space of probability measures over the action space. Next, we formulate a general RL optimization problem and construct a gradient flow using Otto's calculus. We compute the gradient and the Hessian of the energy, providing a formal second-order analysis. Finally, we illustrate the method with numerical examples for low-dimensional problems, computing the gradient directly from our theoretical formalism. For high-dimensional problems, we parameterize the policy using a neural network and optimize it based on an ergodic approximation of the cost.