Edge-Based Data Analytics Solutions logopeterindia.net
Knowledge Hub

Edge-Based Data Analytics Solutions

Platforms, runtimes and frameworks that analyse data where it is created — on gateways, devices and factory floors — for low latency, lower bandwidth use and local control.

39 topics•6 categories•Updated October 2026

Cloud-Extended Edge

5 entries

Cloud providers' edge runtimes and on-premises hardware.

01AWS IoT GreengrassOpen-source edge runtime and cloud service for building, deploying and managing device software, including local processing and machine learning inference. Best for: Running AWS-managed logic and ML on devices. 02Azure IoT OperationsMicrosoft's unified data plane for the edge, which connects, processes and routes operational data close to where it is produced. Best for: Industrial data pipelines in the Azure ecosystem. 03Azure Stack EdgeAzure-managed device that provides compute, storage and intelligence at the edge. Best for: Local analytics and inference with Azure management. 04Google Distributed CloudExtends Google Cloud infrastructure and services to the edge and into your own data centres. Best for: Running Google Cloud services on site or at the edge. 05AWS OutpostsAWS infrastructure and services run on premises. Best for: Keeping AWS APIs and tooling at sites that need local processing.

Industrial & Streaming Edge

7 entries

Industrial edge platforms and data pipelines.

06Siemens Industrial EdgeEdge computing architecture for industrial automation, maintenance, IT-OT connectivity and plant data analytics. Best for: Factory-floor analytics and apps on Siemens automation. 07Litmus EdgeConnects industrial systems, structures operational data and runs analytics and AI at the edge. Best for: Unifying data from mixed industrial equipment. 08ClearBladeIoT, edge AI and digital-twin platform for keeping critical systems efficient and connected. Best for: Edge AI across connected assets and sites. 09CrosserHybrid-first streaming analytics and integration software that runs in the cloud, on premises or at the edge. Best for: Streaming pipelines that span edge and cloud. 10NstreamPlatform for building streaming data applications that react to events as they arrive. Best for: Real-time, stateful applications on live data. 11ThingsBoardOpen-source IoT platform for data collection, processing and dashboards. Best for: Device management with built-in visualisation. 12CumulocityIndustrial AIoT platform for connecting equipment and analysing its data. Best for: Industrial asset monitoring.

Open-Source Edge

7 entries

Open frameworks and databases for edge analytics.

13EdgeX FoundryFlexible, scalable open-source edge platform that supports interoperability between devices and applications at the IoT edge. Best for: Vendor-neutral device connectivity and data collection. 14LF Edge Open HorizonOpen-source project for managing and deploying software workloads to large fleets of edge nodes. Best for: Deploying and updating analytics workloads across many nodes. 15Apache NiFi MiNiFiLightweight agent from the Apache NiFi project for collecting and forwarding data from edge locations. Best for: Collecting data at small edge sites and feeding it upstream. 16InfluxDBTime series database widely used to store and query sensor and telemetry data. Best for: Storing and querying telemetry near the source. 17KubeEdgeExtends Kubernetes to edge nodes, with offline operation and device management. Best for: Running Kubernetes-managed applications at the edge. 18Apache IoTDBTime series database for IoT data with edge and cloud deployment. Best for: Storing and querying high-volume sensor data. 19QuestDBLow-latency time series database. Best for: Fast SQL over time-stamped data on edge servers.

Edge AI & Vision

7 entries

Inference runtimes and vision platforms.

20Edge ImpulsePlatform for building, deploying and monitoring machine learning models across edge hardware. Best for: Training and shipping models to constrained devices. 21NVIDIA MetropolisNVIDIA's vision AI framework for analysing video and sensor data to automate physical spaces and operations. Best for: Video analytics on GPU-equipped edge systems. 22Google LiteRTGoogle's framework, formerly TensorFlow Lite, for running machine learning and generative AI models on device with CPU, GPU and NPU acceleration. Best for: On-device inference on phones and embedded hardware. 23ONNX RuntimeCross-platform engine that accelerates machine learning inference with your existing model frameworks. Best for: Portable, optimised inference across hardware. 24OpenVINOIntel toolkit for optimising and running inference on CPUs, GPUs and NPUs. Best for: Inference on Intel-based edge hardware. 25ExecuTorchPyTorch runtime for on-device AI inference. Best for: Running PyTorch models on phones and embedded devices. 26NVIDIA JetsonEmbedded modules for AI at the edge. Best for: GPU-accelerated vision and robotics workloads.

Fleet & Application Management

3 entries

Deploying and managing software across many edge sites.

27ZEDEDAOrchestration and management of edge devices and applications. Best for: Managing large fleets of edge nodes. 28AvassaEdge application management and operations. Best for: Deploying and running containerised apps across many sites. 29balenaDevice management for deploying and updating containerised software on IoT fleets. Best for: Updating devices in the field.

Related Pages

10 entries

Related directories on this site.

30Edge-Based Streaming Data Analytics SolutionsBrokers, stream processors and collectors for data in motion at the edge. 31Cloud-Based Data Analytics SolutionsThe companion page for analytics in the cloud. 32Edge Computing PlatformsPlatforms for running workloads at the edge. 33Cloud-Native Edge ComputingCloud-native approaches to the edge. 34Edge AIOn-device intelligence. 35Edge Kubernetes PlatformsKubernetes for distributed sites. 36IoT Data Analytics PlatformsPlatforms for sensor and device data. 37IoT and Edge DatabasesData stores that run at the edge. 38Streaming Analytics PlatformsReal-time stream processing. 39Data AnalyticsThe data analytics hub.

Frequently Asked Questions

What is edge-based data analytics?

It is analysing data close to where it is generated, on devices, gateways or local servers, instead of sending everything to a central cloud first. Only results or selected data are then forwarded.

Why analyse data at the edge?

Edge analytics reduces delay, saves network bandwidth, keeps working when connectivity drops and can keep sensitive data on site.

Does edge analytics replace the cloud?

No. Most designs are hybrid: the edge handles fast local decisions and filtering, while the cloud provides long-term storage, large-scale analytics and model training.

What hardware do I need?

It ranges from small gateways and single-board computers to industrial PCs and GPU-equipped servers. Match the hardware to the workload, such as simple threshold rules versus video analytics.

How are edge models kept up to date?

Through fleet management and deployment tools that push new software and models to edge nodes, with monitoring to confirm they are running correctly.

How do I choose an approach?

For decisions in milliseconds, process on the device or gateway. With limited bandwidth, filter and aggregate at the edge. If data must stay on site, analyse locally and share only summaries. For fleets of sites, add an orchestration layer such as Open Horizon or a cloud-managed edge service, and send curated data to the cloud for long-term analysis and model training.