Resources
Frequently Asked Questions
Get answers to common questions about TwinEdge platform capabilities, hardware options, protocol support, ML analytics, operations intelligence features, and pricing.
Platform & Architecture
What is TwinEdge?
TwinEdge is an edge-to-cloud industrial intelligence platform with three integrated components: TwinEdge OS (edge gateway for data acquisition and ML inference), Cloud Platform (fleet management, analytics, and BI Studio), and Operations Intelligence (asset registry, work orders, and GIS mapping). It bridges the gap between operational technology and enterprise systems.
How does edge-to-cloud architecture work?
The edge gateway processes data locally at sub-second latency, running ML inference and physics models without cloud connectivity. Data is synchronized to the cloud for fleet-wide analytics, BI Studio visualizations, and centralized asset management. The system works fully offline and queues data for sync when connectivity returns.
What industries do you support?
TwinEdge is designed for water/wastewater treatment, commercial buildings, healthcare facilities, manufacturing plants, and telecom infrastructure. Our physics models and templates cover pumps, blowers, chillers, generators, rectifiers, and battery banks across these verticals.
Is TwinEdge a SCADA replacement?
No, TwinEdge complements your existing SCADA and PLC infrastructure rather than replacing it. We add machine learning, physics-informed analytics, and cloud-based fleet management on top of your current control systems. Many customers run TwinEdge alongside SCADA for advanced analytics without disrupting proven control logic.
Do I need cloud connectivity to use TwinEdge?
No. The edge gateway runs completely independently with full data acquisition, ML inference, alerting, and local dashboard access. Cloud connectivity adds fleet management, BI Studio, centralized operations intelligence, and multi-site analytics, but is not required for single-site deployments.
What databases does TwinEdge use?
The cloud platform uses a multi-tenant RDBMS with time-series and spatial extensions. The edge gateway uses an embedded database for local storage with automatic data aggregation (1-minute, hourly, daily rollups) and Parquet export for efficient long-term archival.
Hardware & Edge
What hardware do I need?
TwinEdge offers three certified hardware options: Hub Pro ($349, Raspberry Pi CM4-based for indoor use), Hub Industrial ($749, DIN-rail mountable for control panels), and Hub Rugged ($1,099, IP67-rated for outdoor installations). All include pre-configured TwinEdge OS and support for 10+ industrial protocols.
Can I use my own hardware?
Yes. TwinEdge OS runs on any ARM64 or x86_64 Linux device with Docker support and minimum 4GB RAM. We provide installation images and Docker Compose configurations for custom deployments. Community support is available for non-certified hardware.
What is the setup time?
15 minutes for basic deployment. Flash the TwinEdge OS image to an SD card or run our Docker Compose stack, power on the device, and connect to the setup wizard at http://twinedge.local. Network discovery and auto-provisioning features accelerate device onboarding.
How much data can the edge gateway store?
Default retention policy: 7 days of raw data, 30 days of 1-minute aggregates, 365 days of hourly aggregates, and 5 years of daily summaries. All retention periods are configurable. Data is automatically exported to Parquet format for long-term archival and cloud sync.
Does it work offline?
Yes, completely. Full data acquisition from industrial devices, ML inference, physics model calculations, multi-channel alerting (LED, SMS, email, Slack), and local dashboard access all function without internet connectivity. Data queues for cloud sync when connection is restored.
How do OTA updates work?
Over-the-air updates for firmware, ML models, and configuration are delivered via MQTT with automatic rollback on failure. Each update maintains a 24-hour backup. Updates can be scheduled for maintenance windows and deployed to fleets with canary rollout support.
Protocols & Connectivity
What protocols are supported?
TwinEdge supports 10 industrial protocols: OPC UA, Modbus TCP/RTU, EtherNet/IP (CIP), Siemens S7, BACnet/IP, MQTT, Sparkplug B, DNP3, PROFINET RT, and analog 4-20mA input. All protocols are included in every deployment with no per-protocol licensing fees.
How do I connect to existing PLCs and controllers?
Use the network discovery feature to scan your facility network for devices, browse available tags/registers, and select the sensors you want to monitor. TwinEdge connects via native protocols (Modbus, S7, EtherNet/IP, BACnet) without requiring gateways or protocol converters.
Is OPC UA required?
No. The edge gateway connects to your devices using their native protocol (Modbus, S7, BACnet, etc.). Internally, TwinEdge uses OPC UA as a unified namespace to normalize data from all protocols, but your devices do not need to support OPC UA.
Can I connect to BACnet building automation systems?
Yes. Full BACnet/IP support via the BAC0 library enables reading and writing properties, COV (Change of Value) subscriptions, and integration with building management systems from Johnson Controls, Siemens, Honeywell, and other BACnet-compliant vendors.
What about MQTT and Sparkplug B?
TwinEdge includes full MQTT 3.1.1 and 5.0 broker and client support with TLS encryption. Sparkplug B provides IIoT-standard device birth/death/data messages for seamless integration with Ignition SCADA, AWS IoT Core, Azure IoT Hub, and other MQTT-based platforms.
Does TwinEdge support serial protocols like Modbus RTU?
Yes. Modbus RTU over RS-485 is fully supported via USB-to-RS485 adapters. The edge gateway auto-detects serial ports and provides a configuration interface for baud rate, parity, stop bits, and slave address mapping.
Analytics & ML
What ML models can run at the edge?
The edge gateway runs ONNX Runtime for inference with pre-trained models including anomaly detection (Isolation Forest), remaining useful life prediction (Random Forest, XGBoost), and efficiency optimization. All models execute locally with 10-50ms latency on Raspberry Pi hardware.
What are physics models?
Physics models are first-principles equations specific to equipment types (pumps, blowers, generators, batteries). They calculate efficiency, head pressure, power consumption, and other derived metrics from sensor data. TwinEdge combines physics models with ML for physics-informed predictions that are more accurate and explainable than pure data-driven approaches.
How fast is edge inference?
Inference latency is 10-50ms per prediction on Raspberry Pi 5 hardware. Batch mode achieves 500 predictions per second throughput. Models are quantized and optimized for edge deployment without sacrificing accuracy.
Can I train custom ML models?
Yes. Use the twinbox-ml library for feature engineering (time domain, frequency domain, wavelet features) and model training. Export to ONNX format and deploy via OTA update or manual upload. Cloud-based AutoML is available in Enterprise tier for automated model development.
What cloud analytics are available?
Cloud platform includes BI Studio for custom dashboards, fleet-wide analytics, AutoML for automated model training, what-if simulation, pump curve analysis, energy optimization recommendations, and cross-site benchmarking.
How does anomaly detection work?
Edge-based anomaly detection uses Isolation Forest and Autoencoder models trained on normal operating data. The system learns baseline patterns and flags deviations in real-time. Anomaly scores trigger configurable alerts via LED, SMS, email, Slack, or automated work order creation.
Operations Intelligence & Asset Management
Does TwinEdge include maintenance and asset management?
Yes. Built-in Operations Intelligence includes asset registry with hierarchical location trees, work order management with mobile access, preventive maintenance scheduling, QR code-based asset provisioning, and GIS mapping with spatial queries and compliance zones.
How does asset provisioning work?
TwinEdge offers 7 provisioning methods: USB drive auto-import (plug in a config file), QR code scan (printed labels on equipment), CSV bulk import (Excel migration), YAML config files (DevOps workflows), cloud sync (centralized fleet management), device clone (replicate configurations), and network discovery (brownfield installations).
Can I import from my existing CMMS?
Yes. CSV and YAML import support bulk migration from legacy systems. Export your asset data, map fields to TwinEdge schema, and import via API or web interface. Free migration assistance is available for Enterprise tier customers.
Does it support GIS mapping?
Yes. Spatial features include 15 geometry types (points, lines, polygons, multi-geometries), compliance zone tracking, geo-document markers, spatial queries (within, intersects, buffer), layer management, and basemap integration with MapLibre GL JS.
How does work order management work?
Create work orders manually or trigger them automatically from alert rules. Mobile-responsive interface for field technicians. Track time, parts, notes, and photos. Integrate with existing ticketing systems via REST API or webhook notifications.
Can I generate QR codes for equipment?
Yes. Generate QR codes containing asset ID, type, protocol configuration, and sensor mappings. Print labels for equipment tagging. Field technicians scan codes with mobile devices to instantly provision assets in TwinEdge without manual data entry.
Pricing & Getting Started
How much does TwinEdge cost?
Professional tier is designed for single-site deployments. Enterprise tier adds fleet management, BI Studio, and multi-tenant features. Hardware is sold separately (Hub Pro $349, Hub Industrial $749, Hub Rugged $1,099). Contact our sales team for pricing tailored to your deployment size and requirements.
Can I see a demo before purchasing?
Yes. We offer live demos where you can see the platform managing real equipment. Our team will walk you through Professional and Enterprise capabilities tailored to your use case. Hardware pre-orders come with a full refund guarantee anytime before shipment. Contact us to schedule a demo or discussion.
What support is included?
Community tier includes documentation, GitHub issues, and community forums. Professional tier adds <8 hour email response time and quarterly check-ins. Enterprise tier provides <2 hour response with dedicated support engineer, Slack/Teams integration, and custom SLA options.
Can I pre-order hardware?
Yes. All three hardware devices (Hub Pro, Hub Industrial, Hub Rugged) are available for pre-order via Stripe checkout. Business email required. Full refund anytime before shipment. Estimated delivery: Q2 2026 for Hub Pro, Q3 2026 for Industrial and Rugged.
What is the typical deployment timeline?
Edge gateway setup takes 15 minutes for basic configuration. Cloud tenant provisioning is same-day. Full enterprise deployment with custom protocol integration, migration assistance, and training typically takes 2-6 weeks depending on site complexity and existing infrastructure.
Do you offer training and onboarding?
Yes. Professional tier includes self-service documentation and video tutorials. Enterprise tier includes live onboarding sessions, custom training for operations and maintenance teams, and ongoing support for new users. Migration assistance is available for qualifying deployments.
Still Have Questions?
Our team is here to help you understand how TwinEdge can transform your industrial operations. Schedule a demo or reach out to discuss your specific requirements.