Architecture of IUDX Platform Adaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning Authors: M. Yashwanth, G. K. Nayak, A. Singh, Y. Simmhan, A. Chakraborty Federated Learning (FL) is a machine learning paradigm that enables clients to jointly train a global model by aggregating the locally trained models without sharing any local training data. In practice, there can often …
Continue Reading(ADeX) Architecture of IUDX Platform Adaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning Authors: M. Yashwanth, G. K. Nayak, A. Singh, Y. Simmhan, A. Chakraborty Federated Learning (FL) is a machine learning paradigm that enables clients to jointly train a global model by aggregating the locally trained models without sharing any local training data. In practice, there can …
Continue Reading(ADeX) Architecture of IUDX Platform Adaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning Authors: M. Yashwanth, G. K. Nayak, A. Singh, Y. Simmhan, A. Chakraborty Federated Learning (FL) is a machine learning paradigm that enables clients to jointly train a global model by aggregating the locally trained models without sharing any local training data. In practice, there can …
Continue ReadingArchitecture of IUDX Platform Adaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning Authors: M. Yashwanth, G. K. Nayak, A. Singh, Y. Simmhan, A. Chakraborty Federated Learning (FL) is a machine learning paradigm that enables clients to jointly train a global model by aggregating the locally trained models without sharing any local training data. In practice, there can often …
Continue ReadingAdaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning Authors: M. Yashwanth, G. K. Nayak, A. Singh, Y. Simmhan, A. Chakraborty Federated Learning (FL) is a machine learning paradigm that enables clients to jointly train a global model by aggregating the locally trained models without sharing any local training data. In practice, there can often be substantial heterogeneity (e.g., …
Continue Reading
Adaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning Authors: M. Yashwanth, G. K. Nayak, A. Singh, Y. Simmhan, A. Chakraborty Federated Learning (FL) is a machine learning paradigm that enables clients to jointly train a global model by aggregating the locally trained models without sharing any local training data. In practice, there can often be substantial heterogeneity (e.g., …
Continue Reading
Privacy-Preserving Data Quality Assessment for Time-Series IoT Sensors Authors: N. Chakraborty, A. Sharma, J. Dutta. H. D. Kumar This paper proposes a novel framework for automated, objective, and privacy-preserving data quality assessment of time-series data from IoT sensors deployed in smart cities. We leverage custom, autonomously computable metrics that parameterise the temporal performance and adherence to a declarative schema document …
Continue Reading
Automatable Data Quality Dimensions for Data Exchange: Formulation and Application Authors: Debarun Sengupta, Anjula Gurtoo, Minnu Malieckal, Jyotirmoy Dutta Large amounts of data get generated and applied in decision making to improve outcomes. However, quality of the data remains an issue as data gets generated from varied sources, in unspecified formats, and variables vary across different types of data. Identifying …
Continue Reading
Ethics and Fair Use Framework for Privacy Preserving Data Sharing Authors: Bita Afsharinia, Anjula Gurtoo, Jyotirmoy Dutta, Minnu Malieckal The study aims to critically evaluate current privacy-preserving technologies and ethical frameworks in data sharing, identifying gaps and proposing a comprehensive, integrated ethical framework. Existing frameworks often fall short in integrating these two aspects effectively, particularly in the context of emerging …
Continue Reading 
Lead Architect, Smart City Living Lab, IIITH
Chief Program Officer, National Urban Digital Mission
Deputy Director & Chief of Operations, CPDTRI Laboratory Director, ADRC PI, Molecular Immunology, Mazumdar Shaw Medical Foundation (MSMF) Director, Beyond Antibody LLP
Till the late 20th century healthcare has been the forte of biology evading the clutches of mathematics. The new millennium digitised life and sciences therein leading to data-explosion. How digitization and data is propelling healthcare for public good will be discussed with specific examples from the world of diagnostics and therapeutics.
TBD
Director, ICMR-NCDIR
Senior Director, MeitY
Professor, Indian Institute of Science, Co-founder ReneLife, Research Consultant
The Digital Personal Data Protection (DPDP) Act 2023 creates a legal framework for the secured use of personal digital data to eliminate abuse. However, the Act’s provisions appear to have been drafted primarily with the e-commerce websites, personal data-based applications, and other digital customers in mind, not the healthcare community. Healthcare in the digital era requires communicating sensitive patient information over the web with clinicians for responsive advice and consultation. The consequences of such data-sharing involving medical personnel are not clear. Besides, the growth of digital health technologies requires patient data to be accessible anonymously and in de-identified form. Elimination of personal information alone does not guarantee de-identification. Also, to what specific detail one must describe the intended use of data for research and technology development is not articulated. Thus, the gray areas in the Act pose a challenge for the Indian medical ecosystem, which could incur penalties under the DPDP Act 2023.
Chairperson, Arghyam
This topic covers the development of a data-centric public health technology system, with insights from the NCD Portal for the National Non-communicable diseases program. The NCD Portal is used in 31 states and has more than 45 million people on treatment for hypertension and diabetes.
It explores the role of data in four key phases of the technology system: design, platform development, operations, and field adoption. In the design phase, data-driven insights, such as search patterns of frontline workers and medical officers optimize the user interface for speed and efficiency.
Platform development focuses on secure management and rule-based access control of data coming through APIs and data streams, transformed and cleaned, and going to the data warehouse for reporting needs.
Daily operations leverage Application Performance Monitoring (APM) tools and telemetry to identify and resolve performance issues proactively. Field adoption is enhanced through data analytics, guiding weekly reports for program review and feedback loops to prioritize feature development.
Embedding data-centricity throughout the system lifecycle supports large-scale, efficient and secure public health service delivery.
Director, Govt of Telangana, Emerging Technologies, Tech for Social Good
Chair & Professor, Isaac Centre for Public Health, IISc
Head, Product Partnerships, Google
National Program Director, Ease of Doing Business, Project Director, NHEV, President, CPOs of India, & Ex-CAG Member
Scientist G, Centre for Development of Telematics
Chairperson, Centre for Society and Policy, IISc
CEO, DeepMatrix
In this talk, titled “Geospatial Data Democracy: A Technical Perspective,” we will explore how advancements in technology can enable broader access to geospatial data for public good. From open data platforms to cutting-edge tools that lower barriers for non-experts, this session will dive into the technical innovations that have potential to democratizing geospatial information. We will also address the challenges of data privacy, standardization, and infrastructure, and discuss how these obstacles can be overcome. By giving some real-world case studies and future directions, this talk will provide a comprehensive technical outlook on how geospatial data can empower communities, enhance public services, and foster transparent governance.
President, Geospatial Business at MapmyIndia
We see a growing need for geospatial data & solutions for situational awareness, informed decision-making and smart governance. The geo continuum of Capture, Create, Model, Visualize, Analyze and Share involve diverse frameworks and workflows driving variety of applications for Public good. Liberalization of Geospatial sector coupled with a series of Impactful Initiatives triggered the growth & adoption of geospatial data accelerating the country’s digital transformation journey.
Surveyor General of India, Govt of India
IAS, Special Chief Secretary, Department of ITE&C & Department of Industries & Commerce, Govt of Telangana
CTO, Elena Geo Systems Pvt Ltd
Head of Engineering, Modular Open Source Identity Platform (MOSIP)
The quad of Awareness, Law, Policy, and Technology. In securing and safeguarding privacy on a grand scale, we encounter an interplay of four pivotal elements: Awareness, Law, Policy, and Technology. This quadrivium represents not merely a technical or administrative framework but a philosophical foundation that shapes our approach to managing and protecting information in an increasingly complex digital world.
Co-Founder, iSPIRT Foundation
Scientist/Engineer “G” and Group Director, Geosciences Group, National Remote Sensing Centre, ISRO
Vice-Chair of the ACM India Special Interest Group on Computer Science Education
Associate Professor, Indian Institute of Technology, Madras
General Manager, Siemens Healthineers
Competency Head-Data, Analytics & Intelligence, Wipro
Professor, Dept. Comp. & Data Sciences, IISc, Bangalore. Founder & CAiO, Zenteiq AiTech Innovations Private Limited.
Founder & CEO, Sirpi Products & Services Private Limited
Executive Director, Gimcare Hospital & Chairperson, DHIA
Founder and CEO, Garudalytics
This presentation will focus on the application of GeoAI technologies in agriculture and how available data can be utilised to drive innovation and efficiency in farming practices. Emphasising the capabilities of Garudalytics’ smart mapping and AI solutions, we will explore how these tools enhance agricultural decision-making through advanced spatial analysis and predictive modelling. The discussion will include real-world examples where GeoAI has improved crop monitoring, yield prediction, resource management, and overall yield optimization. We will also address the practical challenges of integrating GeoAI into existing agricultural systems and offer strategies for maximising its benefits. Additionally, the talk will cover how GeoAI can leverage both structured and unstructured datasets to enhance agricultural outcomes, with a focus on Garudalytics’ Agri AI solution for integrating diverse datasets, including in-situ data.
CEO, Wadhwani Institute for Artificial Intelligence
We work in the social space, focusing on Applied AI. The space presents unique challenges and constructs that offer an excellent opportunity to have the best minds work through complex problems. The talk focuses on AI in this sector, using a case study-based approach and abstract principles that we present through our doing. This talk is a substrate of our experience in a “learn by doing” approach that we have found to be meaningful.
Dr Brahm Prakash Professor, NRSC/ISRO
Executive Program Director, Sustainable Cities & Director, WRI India Ross Centers
Engineering Director, CDPG
Principal Scientist & Program Manager, CDPG
Engineering Director, CDPG
Technical Lead, Analytics Systems, CDPG
Director, Solvesustain
Founder & Executive Director, Shelter Associates
“Accurate data is the lifeblood of good policies and decision making”
As nations and organisations strive to solve social problems around the world, the lack of accurate data has often been cited as a primary reason why policies do not translate into effective on-ground solutions. The reality in urban informal settlements is no different. They manifest multidimensional deprivation, characterised by a lack of secure tenure, inaccessibility, and inadequate sanitation and health services which marginalises the slum dwellers. Governments and NGOs also find it extremely difficult to effectively target vulnerable communities due to the lack of reliable data. Data on slum settlements is crucial for the inclusion of these underserved communities in city level development plans.
Realising the dire need for safe and hygienic sanitation facilities and practices, Pratima started leveraging data and technology since the late 90s when she realised that data with the Urban Local Bodies was inconclusive and disjointed with the on-ground reality. This propelled the need for a data system that would provide unfailing results. A year of continuous efforts and iterations resulted in a well-tailored Geographical Information System (GIS). It catered to the needs of slum mapping, and spatial layout formation. It also helped to link the socio-economic surveys taken to individual dwellings/house structures that were mapped on GIS. Our granular spatial data harnesses the power of geo-visualisation for effective planning and monitoring of our Sanitation, Housing and Covid-relief projects.
The communities form the backbone of Shelter Associates’ data-driven approach, as they are involved in conducting survey, creating awareness and making sure the data sets are inclusive. This spatial data is then shared with the respective Urban Local Bodies for validation. Once the process is complete, the data is made public on our website while it is also linked to the ULB’s website making it freely available for all to use.
The co-created data helps immensely in bridging the information gap. SA gives slum-dwellers a voice by arming them with data about their community, empowering them to make decisions making them partners in the developmental process.
Founder & Trustee, Oorvani Foundation
OpenCity.in is a leading urban data platform that democratises city data across India by enhancing transparency, fostering understanding and facilitating civic engagement. Alongside hosting an urban open data repository, we work to build a strong open data community.
OpenCity collaborative events or “jams” enable citizens to participate in finding solutions to urban issues; they serve as an effective tool for public problem solving. These events bring together participants from diverse and multidisciplinary backgrounds to open up data, analyse it, and derive actionable insights. From water management to urban mobility, our crowdsourcing projects, datajams, hackathons, and solution jams, have become instrumental in breaking down data silos, revealing patterns, and constructing “what-if” scenarios to propose innovative solutions to pressing public issues.
Through these initiatives, OpenCity provides opportunities for the public to engage with urban issues, which can lead to actionable insights and potential solutions. Our approach brings the power of collective intelligence to drive civic action for better cities.
Research Professor, CiSTUP, IISc
Director, Centre for Brain Research, IISc, Bangalore
The talk will present details of two longitudinal cohort studies in the context of ageing brain research. One study is in the urban area and the other in a rural area. The challenges of sustaining a large cohort of several thousand participants over several years along with the handling of large data that is generated from such studies will be presented.
Road Safety Ambassador, International Road Federation Advisor, Project India Road Safety 2.0, WEF
Director, Engineering & Growth, Namma Yatri
Executive Officer, Head Powertrain & xEV Engineering Maruti Suzuki India Limited
Sr. Operations Officer, World Bank Health Systems
Founder & Director, Data For India
CEO, Health System Transformation Platform
Deputy Surveyor General, NIGST, SOI, Hyderabad
Associate Professor, IISc
Digital interfaces and communications have become critical for access to information, entertainment, economic opportunities and even essential services such as healthcare. Speech is the most natural form of communication, and most Indians are not digital natives comfortable with typing or English. We need spoken language interfaces that work for all variations and dialects. But such data barely exists. IISc-ARTPARK’s umbrella of open-source speech data initiatives are capturing the true diversity of India’s spoken languages to propel language AI technologies and content for an inclusive digital India, in collaboration with Bhashini, MEITY.
TBD
TBD
CEO, ARTPARK, IISc
High-quality data is the key to realizing the promise of AI in healthcare. ICMR (Indian Council of Medical Research), IISc and ARTPARK are building MIDAS India, a national system for gold standard datasets to enable research, validation and deployment of AI-based solutions. It is a DPI (digital public infrastructure)-enabled, thematic hubs-and-spokes system through which purpose-appropriate datasets that reflect the diversity of India’s population, settings and needs can be collected, curated and made available. The effort has been kicked off in collaboration with AIIMS New Delhi and several other institutions across the country, starting with Oral Cancer screening.
General Manager, Strategic Accounts, Invas Technologies (P) Ltd.
SMARTKHETI is a One M2M based integrated platform designed to revolutionise modern farming by providing farmers with comprehensive tools for automation, real-time visibility, and market access. This proposal outlines a cutting-edge system that combines Internet of Things (IoT) technology with web-based applications to offer farmers advanced capabilities, including farm automation, real-time crop monitoring, weather updates, government alerts, market pricing, and equipment rental services. Smart Kheti is certified by ICAR, Delhi
CTO, Genesys International Corporation Ltd
The advent of 3D urban spatial Digital Twins marks a significant milestone in the management of modern cities. This presentation will delve into the transformative potential of these technologies, particularly in the context of India’s National Geospatial Policy 2022. By exploring how Genesys’ content program is spearheading the creation of detailed, city-centric datasets, the talk will highlight the practical applications of these Digital Twins. These datasets not only enhance urban planning and infrastructure management but also foster a new paradigm in how cities are monitored, maintained and developed. Attendees will gain insights into the critical role geospatial data plays in shaping smarter, more resilient urban environments for the future.
CTO, CDPG
As organisations increasingly rely on data-driven insights to fuel innovation, collaboration and decision-making, the need to process sensitive data, such as data containing PII, confidential business information etc., securely has become more critical than ever. Further, the organisations are increasingly tasked with balancing the utility of data with the need to protect individual privacy. Traditional methods of data protection and data anonymization often fall short, especially when computations are performed on datasets containing sensitive information. Privacy Enhancing Technologies (PETs) offer a transformative technological approach, enabling secure computation and further, enabling balancing privacy and utility while working with the underlying data.
This talk will explore the integration of PETs into computational processes involving sensitive and private data, and securing data in its different states: at rest, in transit, in-use and post-use. We will discuss the role of foundational PETs such as Confidential computation with hardware based Secure Enclaves, secure multi-party computation, homomorphic encryption, and federated learning, examining how these technologies protect data across its lifecycle, ensuring that sensitive information remains secure whether it is stored, transmitted, actively processed or released after undergoing some aggregation transformation. We will also discuss implementation, performance and deployment issues concerning PETs. We will present different use cases where secure computation is paramount and PETs become a key enabler. Overall the talk will provide a good overview on how PETs can be effectively incorporated into computational workflows to enhance data privacy across all stages and provide some actionable insights on leveraging PETs to securely manage and compute sensitive data in your organization.
TBD
TBD
Program Director – Ease of Doing Business, Project Director- NHEV, President – CPOS India, Director DIISHA
The panel discussion on “Roadblocks to Achieving Sustainable Mobility” will explore the critical challenges hindering the transition to sustainable transportation systems. Experts from various sectors will examine barriers such as inadequate infrastructure, policy gaps, technological limitations, and societal resistance. The discussion will also focus on innovative solutions and collaborative approaches needed to overcome these obstacles, ensuring a more sustainable and equitable mobility future for all.