AWS IoT SiteWise vs Canary Edge: Which Is Better for Anomaly Detection?

What Is AWS IoT SiteWise Anomaly Detection?

AWS IoT SiteWise added native anomaly detection in July 2025 as the official in-ecosystem replacement for AWS Lookout for Equipment. It provides no-code anomaly monitoring that runs directly on your SiteWise asset models, eliminating the need to build separate ML pipelines.

SiteWise anomaly detection works by analyzing the properties you define in your asset models. You create an asset model (for example, "Industrial Pump"), define properties (flow rate, temperature, vibration), ingest data through SiteWise's data streams, and enable anomaly detection on the model. SiteWise then learns normal operating patterns and flags deviations.

The key requirement is that you must use the SiteWise asset modeling framework. Your equipment must be represented as SiteWise assets with defined properties, measurement streams, and optionally a hierarchy (plant > line > machine > sensor). If you are not already using SiteWise for asset management, adopting anomaly detection means adopting the entire SiteWise platform.

How Does Canary Edge Differ from SiteWise?

Canary Edge is a standalone anomaly detection API with no asset modeling requirement, no framework to adopt, and no cloud lock-in. You send time series JSON data to an HTTP endpoint and receive anomaly scores in the response.

The core architectural difference is dependency scope. SiteWise anomaly detection is a feature within a larger platform — it requires SiteWise asset models, SiteWise data ingestion, SiteWise dashboards, and an AWS account. Canary Edge is a single-purpose API: it does anomaly detection and nothing else.

This difference matters most in three scenarios: when you use multiple cloud providers, when your data pipeline already exists outside AWS, or when you want anomaly detection without re-architecting your equipment data model. Canary Edge accepts data from any source via a standard REST call and returns results in the same HTTP response.

How Do the Features Compare Side by Side?

The two services target different use cases with fundamentally different architectures. Here is a detailed feature comparison.

SiteWise provides a richer monitoring dashboard and integrates natively with other AWS IoT services like IoT Core, IoT Events, and IoT Greengrass. Canary Edge provides a more granular anomaly classification with its 4-level regime system and per-channel contribution scoring for root cause diagnosis.

When Should You Choose SiteWise Over Canary Edge?

Choose SiteWise when your team is fully committed to the AWS IoT ecosystem and you need more than just anomaly detection. SiteWise is the right choice if you already use SiteWise for asset management or are willing to adopt it.

SiteWise is better when: - You already have equipment modeled as SiteWise assets - You need the full IoT platform: data ingestion, asset management, dashboards, and anomaly detection in one service - Your entire infrastructure runs on AWS and you want a single vendor - You prefer no-code configuration over API integration - You need SiteWise's integration with AWS IoT Greengrass for edge computing

Canary Edge is better when: - You want anomaly detection without adopting a new platform - Your data pipeline already exists (Kafka, MQTT broker, custom collectors) - You use multiple cloud providers or run on-premises - You need sub-50ms synchronous API responses for real-time decisions - You want 4-level regime classification instead of binary anomaly flags - You need per-channel contribution scores for root cause analysis - Your team prefers API-first integration over console-based configuration

How Does Setup Time Compare Between the Two?

SiteWise anomaly detection requires 4-8 weeks of setup including asset modeling, while Canary Edge can be producing detection results within hours.

SiteWise setup (4-8 weeks): 1. Week 1-2: Define asset models with properties and measurement definitions 2. Week 2-3: Configure data ingestion (IoT Core rules, SiteWise Edge, or direct API) 3. Week 3-4: Build asset hierarchy (optional but recommended) 4. Week 4-6: Ingest sufficient baseline data for anomaly detection to calibrate 5. Week 6-8: Configure anomaly detection settings, alert thresholds, and dashboards

Canary Edge setup (1-5 days): 1. Day 1: Sign up, generate API key at canaryedge.com 2. Day 1-2: Send healthy operating data to POST /v1/baseline/multivariate (100+ points per channel) 3. Day 2-3: Integrate POST /v1/detect/multivariate into your data pipeline 4. Day 3-5: Test with production data, tune sensitivity

The difference comes down to scope. SiteWise requires you to model your equipment before you can detect anomalies. Canary Edge only needs a sample of healthy data — it learns the rest. AWS published migration scripts on their GitHub (aws-samples repository) to help Lookout customers move to SiteWise, but these scripts handle data format conversion, not the asset modeling work.

Which Is Easier to Migrate to from AWS Lookout for Equipment?

For teams already in the AWS ecosystem with asset models defined, SiteWise is the path of least resistance. For teams that want to decouple from AWS or minimize migration scope, Canary Edge is faster.

Migrating Lookout to SiteWise: - You must create SiteWise asset models for every piece of equipment Lookout monitored - Historical data in Lookout's S3 format must be re-ingested through SiteWise's data pipeline - AWS provides GitHub-hosted migration scripts for data format conversion - Expect 4-8 weeks to be fully operational - You stay within AWS, which simplifies billing and IAM

Migrating Lookout to Canary Edge: - Export healthy data from Lookout's S3 buckets - Reformat CSVs to Canary Edge's JSON channel format - Create baselines via POST /v1/baseline/multivariate — no asset modeling required - Replace Lookout API calls with Canary Edge REST calls - Expect 1-4 weeks to be fully operational - No AWS dependency after migration

The key tradeoff is ecosystem integration vs. simplicity. SiteWise gives you anomaly detection plus a full IoT platform. Canary Edge gives you anomaly detection as a standalone service that works with any infrastructure.