Main Dashboard

The Dashboard is the first screen you see when you open Rela AI. At a glance, it shows how your plant is performing right now: how many assets are operational, how many unresolved alarms exist, and whether maintenance tasks are being completed on time. It is the starting point for any operational decision during a shift.

What is it for?

The Dashboard solves a common problem in industrial plants: information is scattered across different systems and no one has a consolidated view of the operation. A supervisor has to check the SCADA, then email, then the maintenance system, and still doesn't have a complete picture.

The Dashboard centralizes the most important indicators on a single screen, updated every 60 seconds. It allows a shift supervisor, when starting their shift, to know in under a minute whether there is anything urgent to address.

How does it work?

The system collects data from all connected sources (machine agents, tasks, alarms, data sources) and aggregates them into automatically calculated indicators. Each KPI card shows the current value and a 7-day micro-trend so you can see whether the indicator is improving or worsening.

Data refreshes automatically every 60 seconds. If you need the most current status at any moment, use the manual refresh button in the top-right corner.

How to use it?

1. Access the Dashboard from the home icon in the sidebar.
2. Review the four KPI cards at the top to get a general status.
3. If any card shows a red value or a declining trend, click it to go to the detail view.
4. Review the trends chart to see whether events are increasing or decreasing in the selected period.
5. Review the active alarms table to see if there are critical alerts without attention.
6. Use global filters to focus on a specific zone or shift.

Main KPIs

The dashboard displays four key indicators in top cards:

Each card includes a 7-day sparkline and a percentage change indicator compared to the previous period. A green upward arrow on Availability is good; a red arrow on MTTR means repairs are taking longer than before.

Event trends chart

The trend chart shows how many events the system has received in the selected period (24 hours, 7 days, or 30 days), organized by type:

• Alarms — Machine events requiring attention
• Readings — Sensor data and measurements
• Maintenance — Executed maintenance activities
• Agent actions — Automatic responses generated by AI agents

A sudden spike in alarms on a specific day may indicate an unusual event (problematic shift change, extreme environmental condition, cascading failure).

You can click any bar in the chart to see the list of events for that specific period.

Task trends chart

Shows the evolution of created versus completed tasks over time:

If the "created" line consistently exceeds "completed," the team is accumulating a backlog. If the "overdue" line grows, there is an SLA compliance problem to investigate.

Active alarms table

The table shows active alerts sorted from highest to lowest severity:

Critical alarms are highlighted in red and, if you have browser notifications enabled, generate an immediate alert.

Global filters

Filters apply simultaneously to all Dashboard widgets:

• Zone / Location — Focus on a specific area of the plant (Line A, Boiler Room, Packaging Area)
• Department — Show only assets and tasks for the selected department
• Date range — Change the analysis period to view historical trends
• Shift — Filter by operational shift (morning, afternoon, night)

Filters are saved in the page URL, allowing you to copy and share a specific view with a colleague.

Key benefits

• Consolidated view of plant status in under a minute
• Early detection of negative trends (increasing MTTR, accumulating alarms)
• Quick access to the most critical alarms without reviewing multiple systems
• Zone and shift filters so each supervisor sees only what is relevant to them
• 7-day trends showing whether operations are improving or deteriorating

Common use cases

Scenario 1: Start-of-shift review The afternoon supervisor opens Rela AI at 2:00 PM. They review the Dashboard and see 3 active alarms (2 medium, 1 high) and 8 open tasks. The morning shift MTTR was 45 minutes, same as yesterday. Everything normal. They filter by their area (Bottling Line) and see that the "high" alarm is for pump B-12, active for 2 hours. They click it to open the detail and investigate.

Scenario 2: Detecting weekly deterioration The maintenance manager reviews the Dashboard on Monday mornings. This week, MTTR went from 38 to 67 minutes and availability dropped from 97% to 93%. The trend chart shows an alarm spike on Thursday. They switch to a "7 days" period and see that Thursday had 23 events, double the average. They identify that a compressor had multiple cascading failures that day.

Scenario 3: Sharing status with management The supervisor copies the Dashboard URL with the "this week" filter applied and sends it to the plant manager via WhatsApp. The manager opens the link and sees the exact same view with updated KPIs.