arXiv subjects

Yogesh Simmhan

Publications and source records attributed to Yogesh Simmhan.

At least 37 records · Page 2Linked to original sources

arXiv2026

Scaling Real-Time Traffic Analytics on Edge-Cloud Fabrics for City-Scale Camera Networks

Real-time city-scale traffic analytics requires processing 100s-1000s of CCTV streams under strict latency, bandwidth, and compute limits. We present a scalable AI-driven Intelligent Transportation System (AIITS) designed to address multi-dimensional scaling on an edge-cloud fabric. Our platform transforms live multi-camera video feeds into a dynamic traffic graph through a DNN inferencing pipeline, complemented by real-time nowcasting and short-horizon forecasting using Spatio-Temporal GNNs. Using a testbed to validate in a Bengaluru neighborhood, we ingest 100+ RTSP feeds from Raspberry Pis, while Jetson Orin edge accelerators perform high-throughput detection and tracking, producing lightweight flow summaries for cloud-based GNN inference. A capacity-aware scheduler orchestrates load-balancing across heterogeneous devices to sustain real-time performance as stream counts increase. To ensure continuous adaptation, we integrate SAM3 foundation-model assisted labeling and Continuous Federated Learning to update DNN detectors on the edge. Experiments show stable ingestion up to 2000 FPS on Jetson Orins, low-latency aggregation, and accurate and scalable ST-GNN forecasts for up to 1000 streams. A planned live demonstration will scale the full pipeline to 1000 streams, showcasing practical, cross-fabric scalability.

cs.DC↗

arXiv2026

AeroDaaS: A Programmable Drones-as-a-Service Platform for Intelligent Aerial Systems

The increasing adoption of UAVs equipped with advanced sensors and GPU-accelerated edge computing has enabled real-time AI-driven applications in domains such as precision agriculture, wildfire monitoring, and environmental conservation. However, the integrated design and orchestration of navigation, sensing, and analytics, together with seamless real-time coordination across drone, edge, and cloud resources, remains a significant challenge. To address these challenges, we propose AeroDaaS, a service-oriented framework that abstracts UAV-based sensing complexities and provides a Drone-as-a-Service (DaaS) model for intelligent decision-making. AeroDaaS offers modular service primitives for on-demand UAV sensing, navigation and analytics as composable microservices, ensuring cross-platform compatibility and scalability across heterogeneous UAV and edge-cloud infrastructures. AeroDaaS also supports plug-and-play scheduling modules, including Waypoint and Analytics schedulers, which enable trajectory optimization and real-time coordination of inference workloads. We implement and evaluate AeroDaaS for six real-world DaaS applications, of which two are evaluated in real-world scenarios and four in simulation. AeroDaaS requires less than 40 lines of code for the applications and has minimal platform overhead of less than 20 ms per frame and about 1 GB memory usage on Orin Nano and a AMD RTX 3090 GPU workstation. These results are promising for AeroDaaS as an efficient, flexible and scalable UAV programming framework for autonomous aerial analytics.

cs.DC↗

arXiv2026

RIPPLE++: An Incremental Framework for Efficient GNN Inference on Evolving Graphs

Real-world graphs are dynamic, with frequent updates to their structure and features due to evolving vertex and edge properties. These continual changes pose significant challenges for efficient inference in graph neural networks (GNNs). Existing vertex-wise and layer-wise inference approaches are ill-suited for dynamic graphs, as they incur redundant computations, large neighborhood traversals, and high communication costs, especially in distributed settings. Additionally, while sampling-based approaches can be adopted to approximate final layer embeddings, these are often not preferred in critical applications due to their non-determinism. These limitations hinder low-latency inference required in real-time applications. To address this, we propose RIPPLE++, a framework for streaming GNN inference that efficiently and accurately updates embeddings in response to changes in the graph structure or features. RIPPLE++ introduces a generalized incremental programming model that captures the semantics of GNN aggregation functions and incrementally propagates updates to affected neighborhoods. RIPPLE++ accommodates all common graph updates, including vertex/edge addition/deletions and vertex feature updates. RIPPLE++ supports both single-machine and distributed deployments. On a single machine, it achieves up to $56$K updates/sec on sparse graphs like Arxiv ($169$K vertices, $1.2$M edges), and about $7.6$K updates/sec on denser graphs like Products ($2.5$M vertices, $123.7$M edges), with latencies of $0.06$--$960$ms, and outperforming state-of-the-art baselines by $2.2$--$24\times$ on throughput. In distributed settings, RIPPLE++ offers up to $\approx25\times$ higher throughput and $20\times$ lower communication costs compared to recomputing baselines.

cs.DC↗

arXiv2025

Adaptive Self-Distillation for Minimizing Client Drift in Heterogeneous Federated Learning

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., class imbalance) across the local data distributions observed by each of these clients. Under such non-iid label distributions across clients, FL suffers from the 'client-drift' problem where every client drifts to its own local optimum. This results in slower convergence and poor performance of the aggregated model. To address this limitation, we propose a novel regularization technique based on adaptive self-distillation (ASD) for training models on the client side. Our regularization scheme adaptively adjusts to each client's training data based on the global model's prediction entropy and the client-data label distribution. We show in this paper that our proposed regularization (ASD) can be easily integrated atop existing, state-of-the-art FL algorithms, leading to a further boost in the performance of these off-the-shelf methods. We theoretically explain how incorporation of ASD regularizer leads to reduction in client-drift and empirically justify the generalization ability of the trained model. We demonstrate the efficacy of our approach through extensive experiments on multiple real-world benchmarks and show substantial gains in performance when the proposed regularizer is combined with popular FL methods.

cs.LG↗

arXiv2025

SageServe: Optimizing LLM Serving on Cloud Data Centers with Forecast Aware Auto-Scaling

Global cloud service providers handle inference workloads for Large Language Models (LLMs) that span latency-sensitive (e.g., chatbots) and insensitive (e.g., report writing) tasks, resulting in diverse and often conflicting Service Level Agreement (SLA) requirements. Managing such mixed workloads is challenging due to the complexity of the inference serving stack, which encompasses multiple models, GPU hardware, and global data centers. Existing solutions often silo such fast and slow tasks onto separate GPU resource pools with different SLAs, but this leads to significant under-utilization of expensive accelerators due to load mismatch. In this article, we characterize the LLM serving workloads at Microsoft Office 365, one of the largest users of LLMs within Microsoft Azure cloud with over 10 million requests per day, and highlight key observations across workloads in different data center regions and across time. This is one of the first such public studies of Internet-scale LLM workloads. We use these insights to propose SageServe, a comprehensive LLM serving framework that dynamically adapts to workload demands using multi-timescale control knobs. It combines short-term request routing to data centers with long-term scaling of GPU VMs and model placement with higher lead times, and co-optimizes the routing and resource allocation problem using a traffic forecast model and an Integer Linear Programming (ILP) solution. We evaluate SageServe through real runs and realistic simulations on 10 million production requests across three regions and four open-source models. We achieve up to 25% savings in GPU-hours compared to the current baseline deployment and reduce GPU-hour wastage due to inefficient auto-scaling by 80%, resulting in a potential monthly cost savings of up to $2.5 million, while maintaining tail latency and meeting SLAs.

cs.DC↗

arXiv2025

The Urban Vision Hackathon Dataset and Models: Towards Image Annotations and Accurate Vision Models for Indian Traffic

This report describes the UVH-26 dataset, the first public release by AIM@IISc of a large-scale dataset of annotated traffic-camera images from India. The dataset comprises 26,646 high-resolution (1080p) images sampled from 2800 Bengaluru's Safe-City CCTV cameras over a 4-week period, and subsequently annotated through a crowdsourced hackathon involving 565 college students from across India. In total, 1.8 million bounding boxes were labeled across 14 vehicle classes specific to India: Cycle, 2-Wheeler (Motorcycle), 3-Wheeler (Auto-rickshaw), LCV (Light Commercial Vehicles), Van, Tempo-traveller, Hatchback, Sedan, SUV, MUV, Mini-bus, Bus, Truck and Other. Of these, 283k-316k consensus ground truth bounding boxes and labels were derived for distinct objects in the 26k images using Majority Voting and STAPLE algorithms. Further, we train multiple contemporary detectors, including YOLO11-S/X, RT-DETR-S/X, and DAMO-YOLO-T/L using these datasets, and report accuracy based on mAP50, mAP75 and mAP50:95. Models trained on UVH-26 achieve 8.4-31.5% improvements in mAP50:95 over equivalent baseline models trained on COCO dataset, with RT-DETR-X showing the best performance at 0.67 (mAP50:95) as compared to 0.40 for COCO-trained weights for common classes (Car, Bus, and Truck). This demonstrates the benefits of domain-specific training data for Indian traffic scenarios. The release package provides the 26k images with consensus annotations based on Majority Voting (UVH-26-MV) and STAPLE (UVH-26-ST) and the 6 fine-tuned YOLO and DETR models on each of these datasets. By capturing the heterogeneity of Indian urban mobility directly from operational traffic-camera streams, UVH-26 addresses a critical gap in existing global benchmarks, and offers a foundation for advancing detection, classification, and deployment of intelligent transportation systems in emerging nations with complex traffic conditions.

cs.CV↗

arXiv2025

AeroResQ: Edge-Accelerated UAV Framework for Scalable, Resilient and Collaborative Escape Route Planning in Wildfire Scenarios

Drone fleets equipped with onboard cameras, computer vision, and Deep Neural Network (DNN) models present a powerful paradigm for real-time spatio-temporal decision-making. In wildfire response, such drones play a pivotal role in monitoring fire dynamics, supporting firefighter coordination, and facilitating safe evacuation. In this paper, we introduce AeroResQ, an edge-accelerated UAV framework designed for scalable, resilient, and collaborative escape route planning during wildfire scenarios. AeroResQ adopts a multi-layer orchestration architecture comprising service drones (SDs) and coordinator drones (CDs), each performing specialized roles. SDs survey fire-affected areas, detect stranded individuals using onboard edge accelerators running fire detection and human pose identification DNN models, and issue requests for assistance. CDs, equipped with lightweight data stores such as Apache IoTDB, dynamically generate optimal ground escape routes and monitor firefighter movements along these routes. The framework proposes a collaborative path-planning approach based on a weighted A* search algorithm, where CDs compute context-aware escape paths. AeroResQ further incorporates intelligent load-balancing and resilience mechanisms: CD failures trigger automated data redistribution across IoTDB replicas, while SD failures initiate geo-fenced re-partitioning and reassignment of spatial workloads to operational SDs. We evaluate AeroResQ using realistic wildfire emulated setup modeled on recent Southern California wildfires. Experimental results demonstrate that AeroResQ achieves a nominal end-to-end latency of <=500ms, much below the 2s request interval, while maintaining over 98% successful task reassignment and completion, underscoring its feasibility for real-time, on-field deployment in emergency response and firefighter safety operations.

cs.DC↗

arXiv2025

Towards AI Agents for Course Instruction in Higher Education: Early Experiences from the Field

This article presents early findings from designing, deploying and evaluating an AI-based educational agent deployed as the primary instructor in a graduate-level Cloud Computing course at IISc. We detail the design of a Large Language Model (LLM)-driven Instructor Agent, and introduce a pedagogical framework that integrates the Instructor Agent into the course workflow for actively interacting with the students for content delivery, supplemented by the human instructor to offer the course structure and undertake question--answer sessions. We also propose an analytical framework that evaluates the Agent--Student interaction transcripts using interpretable engagement metrics of topic coverage, topic depth and turn-level elaboration. We report early experiences on how students interact with the Agent to explore concepts, clarify doubts and sustain inquiry-driven dialogue during live classroom sessions. We also report preliminary analysis on our evaluation metrics applied across two successive instructional modules that reveals patterns of engagement evolution, transitioning from broad conceptual exploration to deeper, focused inquiry. These demonstrate how structured integration of conversational AI agents can foster reflective learning, offer a reproducible methodology for studying engagement in authentic classroom settings, and support scalable, high-quality higher education.

cs.CY↗

arXiv2025

NeoARCADE: Robust Calibration for Distance Estimation to Support Assistive Drones for the Visually Impaired

Autonomous navigation by drones using onboard sensors, combined with deep learning and computer vision algorithms, is impacting a number of domains. We examine the use of drones to autonomously follow and assist Visually Impaired People (VIPs) in navigating urban environments. Estimating the absolute distance between the drone and the VIP, and to nearby objects, is essential to design obstacle avoidance algorithms. Here, we present NeoARCADE (Neo), which uses depth maps over monocular video feeds, common in consumer drones, to estimate absolute distances to the VIP and obstacles. Neo proposes robust calibration technique based on depth score normalization and coefficient estimations to translate relative distances from depth map to absolute ones. It further develops a dynamic recalibration method that can adapt to changing scenarios. We also develop two baseline models, Regression and Geometric, and compare Neo with SOTA depth map approaches and the baselines. We provide detailed evaluations to validate their robustness and generalizability for distance estimation to VIPs and other obstacles in diverse and dynamic conditions, using datasets collected in a campus environment. Neo predicts distances to VIP with an error <30cm, and to different obstacles like cars and bicycles within a maximum error of 60cm, which are better than the baselines. Neo also clearly out-performs SOTA depth map methods, reporting errors up to 5.3-14.6x lower.

cs.RO↗

arXiv2025

OptimES: Optimizing Federated Learning Using Remote Embeddings for Graph Neural Networks

Graph Neural Networks (GNNs) have experienced rapid advancements in recent years due to their ability to learn meaningful representations from graph data structures. However, in most real-world settings, such as financial transaction networks and healthcare networks, this data is localized to different data owners and cannot be aggregated due to privacy concerns. Federated Learning (FL) has emerged as a viable machine learning approach for training a shared model that iteratively aggregates local models trained on decentralized data. This addresses privacy concerns while leveraging parallelism. State-of-the-art methods enhance the privacy-respecting convergence accuracy of federated GNN training by sharing remote embeddings of boundary vertices through a server (EmbC). However, they are limited by diminished performance due to large communication costs. In this article, we propose OptimES, an optimized federated GNN training framework that employs remote neighbourhood pruning, overlapping the push of embeddings to the server with local training, and dynamic pulling of embeddings to reduce network costs and training time. We perform a rigorous evaluation of these strategies for four common graph datasets with up to $111M$ vertices and $1.8B$ edges. We see that a modest drop in per-round accuracy due to the preemptive push of embeddings is out-stripped by the reduction in per-round training time for large and dense graphs like Reddit and Products, converging up to $\approx 3.5\times$ faster than EmbC and giving up to $\approx16\%$ better accuracy than the default federated GNN learning. While accuracy improvements over default federated GNNs are modest for sparser graphs like Arxiv and Papers, they achieve the target accuracy about $\approx11\times$ faster than EmbC.

cs.DC↗

arXiv2025

Pagoda: An Energy and Time Roofline Study for DNN Workloads on Edge Accelerators

Edge accelerators such as Nvidia Jetsons are becoming an integral part of the computing continuum, and are often used for DNN inferencing and training. Nvidia Jetson edge devices have $2000$+ CUDA cores within a $70$W power envelope and offer $1000$s of power modes to customize CPU, GPU and memory frequencies. Their widely varying power--performance trade-offs can be exploited for energy and power-constrained deployments. While data-driven methods to predict the power and latency of DNN workloads for edge devices exist, there is a lack of principled study to understand why edge accelerators and their power modes perform the way they do. We develop a time roofline and a novel energy roofline model for the Jetson Orin AGX for diverse power modes, and couple it with an analytical model of the compute (FLOP) and memory access (bytes) for DNN inference workloads to analyze them from first principles. These reveal unique, sometimes counter-intuitive, insights into the power and performance behavior of DNN workloads on edge accelerators, e.g., the default power mode MAXN is not the most energy efficient and time efficiency implies energy efficiency for all power modes. We also extend our analytical roofline models to DNN training. Finally, we apply these methods to tune the power mode (and hence the roofline) of the edge device to optimize the latency and energy for DNN inference, with up to $15\%$ lower energy and minimal degradation in inference time.

cs.DC↗

arXiv2025

Fulcrum: Optimizing Concurrent DNN Training and Inferencing on Edge Accelerators

The proliferation of GPU accelerated edge devices like Nvidia Jetsons and the rise in privacy concerns are placing an emphasis on concurrent DNN training and inferencing on edge devices. Inference and training have different computing and QoS goals. But edge accelerators like Jetson do not support native GPU sharing and expose 1000s of power modes. This requires careful time-sharing of concurrent workloads to meet power--performance goals, while limiting costly profiling. In this paper, we design an intelligent time-slicing approach for concurrent DNN training and inferencing on Jetsons. We formulate an optimization problem to interleave training and inferencing minibatches, and decide the device power mode and inference minibatch size, while maximizing the training throughput and staying within latency and power budgets, with modest profiling costs. We propose GMD, an efficient multi-dimensional gradient descent search which profiles just $15$ power modes; and ALS, an Active Learning technique which identifies reusable Pareto-optimal power modes, but profiles $50$--$150$ power modes. We evaluate these within our Fulcrum scheduler for $273,000+$ configurations across $15$ DNN workloads. We also evaluate our strategies on dynamic arrival inference and concurrent inferences. ALS and GMD outperform simpler and more complex baselines with larger-scale profiling. Their solutions satisfy the latency and power budget for $>97\%$ of our runs, and on average are within $7\%$ of the optimal throughput.

cs.DC↗

arXiv2025

Characterizing the Performance of Accelerated Jetson Edge Devices for Training Deep Learning Models

Deep Neural Networks (DNNs) have had a significant impact on domains like autonomous vehicles and smart cities through low-latency inferencing on edge computing devices close to the data source. However, DNN training on the edge is poorly explored. Techniques like federated learning and the growing capacity of GPU-accelerated edge devices like NVIDIA Jetson motivate the need for a holistic characterization of DNN training on the edge. Training DNNs is resource-intensive and can stress an edge's GPU, CPU, memory and storage capacities. Edge devices also have different resources compared to workstations and servers, such as slower shared memory and diverse storage media. Here, we perform a principled study of DNN training on individual devices of three contemporary Jetson device types: AGX Xavier, Xavier NX and Nano for three diverse DNN model--dataset combinations. We vary device and training parameters such as I/O pipelining and parallelism, storage media, mini-batch sizes and power modes, and examine their effect on CPU and GPU utilization, fetch stalls, training time, energy usage, and variability. Our analysis exposes several resource inter-dependencies and counter-intuitive insights, while also helping quantify known wisdom. Our rigorous study can help tune the training performance on the edge, trade-off time and energy usage on constrained devices, and even select an ideal edge hardware for a DNN workload, and, in future, extend to federated learning too. As an illustration, we use these results to build a simple model to predict the training time and energy per epoch for any given DNN across different power modes, with minimal additional profiling.

cs.DC↗

arXiv2025

AgentX: Towards Orchestrating Robust Agentic Workflow Patterns with FaaS-hosted MCP Services

Generative Artificial Intelligence (GenAI) has rapidly transformed various fields including code generation, text summarization, image generation and so on. Agentic AI is a recent evolution that further advances this by coupling the decision making and generative capabilities of LLMs with actions that can be performed using tools. While seemingly powerful, Agentic systems often struggle when faced with numerous tools, complex multi-step tasks,and long-context management to track history and avoid hallucinations. Workflow patterns such as Chain-of-Thought (CoT) and ReAct help address this. Here, we define a novel agentic workflow pattern, AgentX, composed of stage designer, planner, and executor agents that is competitive or better than the state-of-the-art agentic patterns. We also leverage Model Context Protocol (MCP) tools, and propose two alternative approaches for deploying MCP servers as cloud Functions as a Service (FaaS). We empirically evaluate the success rate, latency and cost for AgentX and two contemporary agentic patterns, ReAct and Magentic One, using these the FaaS and local MCP server alternatives for three practical applications. This highlights the opportunities and challenges of designing and deploying agentic workflows.

cs.DC↗

arXiv2025

Optimizing Federated Learning for Scalable Power-demand Forecasting in Microgrids

Real-time monitoring of power consumption in cities and micro-grids through the Internet of Things (IoT) can help forecast future demand and optimize grid operations. But moving all consumer-level usage data to the cloud for predictions and analysis at fine time scales can expose activity patterns. Federated Learning~(FL) is a privacy-sensitive collaborative DNN training approach that retains data on edge devices, trains the models on private data locally, and aggregates the local models in the cloud. But key challenges exist: (i) clients can have non-independently identically distributed~(non-IID) data, and (ii) the learning should be computationally cheap while scaling to 1000s of (unseen) clients. In this paper, we develop and evaluate several optimizations to FL training across edge and cloud for time-series demand forecasting in micro-grids and city-scale utilities using DNNs to achieve a high prediction accuracy while minimizing the training cost. We showcase the benefit of using exponentially weighted loss while training and show that it further improves the prediction of the final model. Finally, we evaluate these strategies by validating over 1000s of clients for three states in the US from the OpenEIA corpus, and performing FL both in a pseudo-distributed setting and a Pi edge cluster. The results highlight the benefits of the proposed methods over baselines like ARIMA and DNNs trained for individual consumers, which are not scalable.

cs.DC↗

arXiv2025

AerialDB: A Federated Peer-to-Peer Spatio-temporal Edge Datastore for Drone Fleets

Recent years have seen an unprecedented growth in research that leverages the newest computing paradigm of Internet of Drones, comprising a fleet of connected Unmanned Aerial Vehicles (UAVs) used for a wide range of tasks such as monitoring and analytics in highly mobile and changing environments characteristic of disaster regions. Given that the typical data (i.e., videos and images) collected by the fleet of UAVs deployed in such scenarios can be considerably larger than what the onboard computers can process, the UAVs need to offload their data in real-time to the edge and the cloud for further processing. To that end, we present the design of AerialDB - a lightweight decentralized data storage and query system that can store and process time series data on a multi-UAV system comprising: A) a fleet of hundreds of UAVs fitted with onboard computers, and B) ground-based edge servers connected through a cellular link. Leveraging lightweight techniques for content-based replica placement and indexing of shards, AerialDB has been optimized for efficient processing of different possible combinations of typical spatial and temporal queries performed by real-world disaster management applications. Using containerized deployment spanning up to 400 drones and 80 edges, we demonstrate that AerialDB is able to scale efficiently while providing near real-time performance with different realistic workloads. Further, AerialDB comprises a decentralized and locality-aware distributed execution engine which provides graceful degradation of performance upon edge failures with relatively low latency while processing large spatio-temporal data. AerialDB exhibits comparable insertion performance and 100 times improvement in query performance against state-of-the-art baseline. Moreover, it exhibits a 10 times and 100 times improvement with insertion and query workloads respectively over the cloud baseline.

cs.DC↗

arXiv2025

D3FL: Data Distribution and Detrending for Robust Federated Learning in Non-linear Time-series Data

With advancements in computing and communication technologies, the Internet of Things (IoT) has seen significant growth. IoT devices typically collect data from various sensors, such as temperature, humidity, and energy meters. Much of this data is temporal in nature. Traditionally, data from IoT devices is centralized for analysis, but this approach introduces delays and increased communication costs. Federated learning (FL) has emerged as an effective alternative, allowing for model training across distributed devices without the need to centralize data. In many applications, such as smart home energy and environmental monitoring, the data collected by IoT devices across different locations can exhibit significant variation in trends and seasonal patterns. Accurately forecasting such non-stationary, non-linear time-series data is crucial for applications like energy consumption estimation and weather forecasting. However, these data variations can severely impact prediction accuracy. The key contributions of this paper are: (1) Investigating how non-linear, non-stationary time-series data distributions, like generalized extreme value (gen-extreme) and log norm distributions, affect FL performance. (2) Analyzing how different detrending techniques for non-linear time-series data influence the forecasting model's performance in a FL setup. We generated several synthetic time-series datasets using non-linear data distributions and trained an LSTM-based forecasting model using both centralized and FL approaches. Additionally, we evaluated the impact of detrending on real-world datasets with non-linear time-series data distributions. Our experimental results show that: (1) FL performs worse than centralized approaches when dealing with non-linear data distributions. (2) The use of appropriate detrending techniques improves FL performance, reducing loss across different data distributions.

cs.LG↗

arXiv2025

Flotilla: A scalable, modular and resilient federated learning framework for heterogeneous resources

With the recent improvements in mobile and edge computing and rising concerns of data privacy, Federated Learning(FL) has rapidly gained popularity as a privacy-preserving, distributed machine learning methodology. Several FL frameworks have been built for testing novel FL strategies. However, most focus on validating the learning aspects of FL through pseudo-distributed simulation but not for deploying on real edge hardware in a distributed manner to meaningfully evaluate the federated aspects from a systems perspective. Current frameworks are also inherently not designed to support asynchronous aggregation, which is gaining popularity, and have limited resilience to client and server failures. We introduce Flotilla, a scalable and lightweight FL framework. It adopts a ``user-first'' modular design to help rapidly compose various synchronous and asynchronous FL strategies while being agnostic to the DNN architecture. It uses stateless clients and a server design that separates out the session state, which are periodically or incrementally checkpointed. We demonstrate the modularity of Flotilla by evaluating five different FL strategies for training five DNN models. We also evaluate the client and server-side fault tolerance on 200+ clients, and showcase its ability to rapidly failover within seconds. Finally, we show that Flotilla's resource usage on Raspberry Pis and Nvidia Jetson edge accelerators are comparable to or better than three state-of-the-art FL frameworks, Flower, OpenFL and FedML. It also scales significantly better compared to Flower for 1000+ clients. This positions Flotilla as a competitive candidate to build novel FL strategies on, compare them uniformly, rapidly deploy them, and perform systems research and optimizations.

cs.DC↗