The bottom line: Most manufacturing software tells you what happened yesterday. ProAlert tells you what's happening now. That's the difference between fixing today's problems today and reading about them in tomorrow's report. The SignalR engine is the reason a downtime call placed on the floor reaches every qualified responder's phone in under one second... and the reason RHA Manufacturing cut response time from 22 minutes to under 1 minute.

Why "Refresh to Update" Is the Wrong Architecture

Polling introduces latency that compounds into lost production.

Traditional manufacturing dashboards poll the server every 30, 60, or 120 seconds. A machine that goes down at second 1 of a 60-second poll cycle stays invisible on every dashboard and mobile device for up to a minute. At 1,000 cycles per hour, that's 1,000 lost units before anyone knows there's a problem.
Broadcast-to-everyone creates noise that everyone learns to ignore.

Systems that push alerts to all users with no structural routing train maintenance teams to ignore notifications. When every alert fires for every asset in the building regardless of role, the signal-to-noise ratio collapses and critical alerts get dismissed with the routine ones.
IoT telemetry mixed into the main alert stream degrades performance.

EdgeSense IoT (Internet of Things) devices transmit continuous telemetry: cycle counts, sensor readings, device heartbeats. Routing that data through the same channel as operator alerts saturates the hub and increases latency for the alerts that matter most.
Disconnected devices have no feedback on what they missed.

When a mobile device goes offline and comes back, most systems simply show current state with no indication of what changed while the connection was down. Operations running from incomplete information make decisions that compound the original problem.

The SignalR Engine: Two Hubs, One Purpose

ProAlert runs two dedicated SignalR hubs to keep operational alerts and IoT device telemetry on separate channels... preventing edge device traffic from ever affecting alert delivery latency.

ProAlertHub (Main Alert Hub)

Central SignalR hub for all operator-facing communication. Downtime calls, OEE metrics, production warnings, video ready notifications, shift events, and asset messaging all route through ProAlertHub. Delivers bidirectional communication to web dashboards, mobile app, and Andon boards simultaneously.

EdgeSenseHub (IoT Device Hub)

Dedicated SignalR hub for IoT edge device communication. Cycle counts, sensor readings, device heartbeats, and offline-sync events route through EdgeSenseHub... keeping high-frequency telemetry completely separate from the main alert stream to protect alert delivery performance.

Sub-Second Delivery

WebSocket persistent connections eliminate the poll-wait-receive latency of HTTP-based systems. From the moment a DT (Downtime) call is initiated to the moment it reaches every qualified responder's device: under one second. No polling interval, no batch job, no delay.

Bidirectional Communication

SignalR connections are bidirectional. The server pushes to clients; clients can push back through the same connection. Call claims, status acknowledgements, and device commands flow over the same persistent WebSocket channel without a separate HTTP round trip.

Seven Notification Types — One Delivery Channel

Every notification type in ProAlert routes through the same sub-second SignalR infrastructure. There is no separate push system, no third-party notification service, and no polling fallback.

Notification Type	Trigger Event	Recipients
Call Notifications	DT call initiated by operator, Andon button, or EdgeSense device	All users in the qualifying BOBR (Bill of Business Resources) group for the affected asset... filtered by role
Live OEE Broadcasts	Every 10 seconds, continuous	All connected clients subscribed to any asset in the system... web dashboards, mobile, Andon boards
Production Warning Alerts	Slow cycle detection, production lull threshold exceeded, or cycles detected while asset is in Down state	Supervisors and operators assigned to the affected asset and its organizational group
Video Ready Notifications	VideoClipProcessingService completes clip export and thumbnail generation	User who initiated the DT call and any users subscribed to the asset
Device Status Updates	EdgeSense device connects, loses connection, or completes offline batch sync	Administrators and supervisors with device management permissions
Shift Event Broadcasting	Shift start, shift stop, break start, break end, and production run transitions	All operators and supervisors assigned to the affected facility and shift
Asset Messaging	Operator or support user sends a peer-to-peer message through the asset messaging interface	Target user or asset group as specified by the sender

BOBR Structural Routing: Right People, Right Alert

ProAlert's BOBR (Bill of Business Resources) hierarchy is the organizational backbone behind every notification. Every alert is targeted to the appropriate group... not broadcast to everyone and trusted to be ignored by the wrong people.

Role-Based Group Membership

Users are assigned to BOBR groups by role: Maintenance, Supervisor, Quality, Operator. A DT call on a press in Building 2 routes to Maintenance users in Building 2... not to Quality users in Building 1.

Multi-Language Support

BOBR group routing includes language preference. Alert content is delivered in the recipient's configured language. Four-language support (English, Spanish, French, Portuguese) for plant floors with multilingual workforces.

Global Multi-Plant Subscription

Support users and corporate administrators subscribe to all calls system-wide (brId=0) for unified visibility across every facility. One dashboard, every plant, every active alert.

Asset-Level Precision

Routing resolves to the individual asset level. A call on Press 4 in Cell B notifies the responders responsible for that specific asset... not everyone in the building. First-Responder-Wins logic then ensures only one person holds the claim.

What Sub-Second Delivery Means in Practice

<1s

Alert-to-device delivery time

82%

Reduction in response time (RHA: 22 min to <1 min)

10s

OEE broadcast interval to all connected screens

131K+

DT calls processed through the alert engine

At RHA Manufacturing, a Toyota Tier 1 (T1) stamping supplier, average response time to downtime events dropped from 22 minutes to under 1 minute after implementing ProAlert's real-time alert system. That reduction in response time was the single largest driver of their OEE climb from 65% to 89% over 39 months... because a problem that gets responded to in under a minute doesn't become a problem that compounds for 22 minutes.

ProAlert vs. Polling-Based Manufacturing Software

Polling-Based Manufacturing Platforms

Dashboard updates every 30, 60, or 120 seconds
Mobile push via third-party notification service with variable delivery delays
Broadcast alerts to all users regardless of role or location
Single notification channel for all event types
No real-time device status or offline-sync visibility
OEE metrics visible only after next scheduled report run

ProAlert Real-Time Intelligence

WebSocket persistent connection — no polling, no latency floor
Native SignalR push to web, mobile, and Andon boards simultaneously
BOBR structural routing targets the right role at the right asset
Two dedicated hubs keep IoT telemetry separate from operator alerts
Device online, offline, and syncing status visible in real time
OEE broadcast to every connected screen every 10 seconds

For IT and Infrastructure Teams

Component	Technology	Notes
Real-Time Transport	ASP.NET Core SignalR (WebSockets, fallback to Server-Sent Events)	WebSockets preferred for sub-second performance. Automatic fallback to Server-Sent Events (SSE) and long polling for environments with WebSocket restrictions. No separate messaging broker required.
Hub Architecture	ProAlertHub + EdgeSenseHub (dedicated IoT hub)	Two independently scaled SignalR hubs. ProAlertHub handles all operator-facing communication. EdgeSenseHub handles edge device telemetry. Separation prevents IoT traffic volume from impacting alert delivery latency.
Group Management	BOBR-keyed SignalR groups, managed in SQL Server	Users are added to and removed from SignalR groups based on their BOBR assignment in the database. Group membership updates in real time when users are reassigned without requiring reconnection.
OEE Broadcast Service	Background worker, 10-second interval broadcast	Dedicated background service reads current OEE metrics from the SQL Server (Structured Query Language Server) view and broadcasts to all connected clients on a fixed 10-second interval. Runs on the same server as the web application.
Network Requirements	WebSocket (port 443 HTTPS) from client to server	Standard HTTPS/WSS (WebSocket Secure) on port 443. No additional firewall rules required beyond standard HTTPS. Compatible with corporate proxies that support WebSocket upgrade. Mobile clients connect over the same endpoint.
Client Libraries	@microsoft/signalr (JavaScript), Microsoft.AspNetCore.SignalR.Client (.NET MAUI)	Web clients use the official SignalR JavaScript library included in the ProAlert bundle. The .NET MAUI (Multi-platform App UI) mobile app uses the native SignalR.Client package for iOS and Android. No additional dependencies required.

See a downtime alert travel from floor to phone in under one second.

Book a 30-minute demo... we'll place a live DT call, watch it arrive on mobile in real time, and walk through the BOBR routing that put it in the right hands.

Schedule a Demo