Dashboard

🚧 Under Development: The HORUS Dashboard is currently under active development. Features and APIs may change. Check back for updates!

Your mission control for robotics applications. Monitor performance, debug issues, and tune parameters - all from your browser, accessible from any device on your network.

What is it?

The HORUS Dashboard is a web interface that lets you see inside your running robot system. Instead of tailing logs or SSH'ing into your robot, you get a visual, real-time view of everything happening.

Access it from:

• Your laptop while the robot runs
• Your phone/tablet for mobile monitoring
• Another team member's computer
• Anywhere on your network

Why use it?

You want to know: "Is my robot working?"

The dashboard shows you at a glance:

• All nodes running? (System health badge)
• Performance good? (CPU, memory, tick rates)
• Messages flowing? (Topic activity)
• Any errors? (Error counts)

Without the dashboard: Grep logs, check processes, SSH into robot, hope you find the issue

With the dashboard: Open browser, see everything instantly

You want to: "Debug why it's not working"

Common problems the dashboard solves:

"My subscriber isn't getting messages"

• Check Topics tab Is the topic there?
• Check Graph view Are publisher and subscriber connected?
• Check Logs Any connection errors?

"The robot is running slow"

• Check Nodes tab Which node is using 90% CPU?
• Check tick rates Which node can't keep up?
• Check Logs Any "slow tick" warnings?

"It worked yesterday, what changed?"

• Check Parameters What values are different?
• Check Nodes Same nodes running as before?

You want to: "Tune parameters live"

Example: Tuning PID controller gains

Without dashboard:

1. Edit config file
2. Stop robot
3. Restart robot
4. Test behavior
5. Repeat 20 times

With dashboard:

1. Open Parameters tab
2. Edit pid.kp value
3. Watch robot respond instantly
4. Tweak until perfect
5. Export final values

Result: 5 minutes instead of an hour

Quick Start

# Start your robot
horus run

# In another terminal, start dashboard
horus dashboard

Browser opens automatically to http://localhost:3000

What You Can Do

1. Monitor Performance

See at a glance:

• All running nodes in a table
• CPU and memory usage per node
• How fast each node is running (tick rate)
• Error counts

When to check:

• Starting up (verify all nodes launched)
• During operation (watch for anomalies)
• After code changes (ensure no performance regression)

What you'll notice:

• Tick rate steady at ~60 Hz Good
• CPU jumping to 90% Check that node
• Errors climbing Something's wrong

2. Debug Message Flow

Visual graph showing:

• Blue circles = Your nodes
• Green circles = Topics
• Arrows = Who's talking to whom

Use it when:

• "Why isn't my subscriber getting data?"

  • Look: Is there an arrow from publisher topic subscriber?
  • If not: Name mismatch or node not running

• "Which nodes use this topic?"

  • Click the topic circle
  • See all connected nodes highlighted

3. Watch Real-Time Logs

See logs from:

• All nodes combined (system-wide view)
• One specific node (filter dropdown)
• Messages on a specific topic

Use it when:

• Debugging errors
• Verifying messages are being sent
• Checking timing (IPC overhead shown)

Better than terminal logs:

• Color-coded by level (info/warning/error)
• Filter by node without grep
• Auto-scrolls with new messages
• Shows performance metrics

4. Tune Parameters Live

Interactive parameter editor:

• See all current values
• Edit any parameter
• Changes apply immediately
• Auto-saves to disk

Perfect for:

• PID tuning (adjust gains while watching behavior)
• Speed limits (test different max velocities)
• Sensor thresholds (tune detection sensitivity)

You'll save hours compared to edit-restart-test cycles

5. Manage Packages

Browse and install:

• Search the registry for packages
• One-click install
• Choose global or local installation

Like npm/pip but for robot components

6. Deploy to Robots

Send code to physical hardware:

• Enter robot's IP/ID
• Select file to deploy
• Click deploy
• Watch build progress
• View robot's logs and stats

Use it when:

• Testing on real hardware
• Running experiments remotely
• Managing multiple robots

Common Scenarios

Starting Your Day

You just sat down with coffee:

1. Open dashboard
2. Check health badge 🟢 All good
3. Glance at nodes table All running, tick rates normal
4. Start working

Total time: 10 seconds to verify system health

Something Broke

Your robot stopped moving:

Dashboard workflow:

1. Health badge is 🔴 Critical
2. Nodes tab motor_controller has 50 errors
3. Click logs for that node
4. See: "Failed to connect to motor driver"
5. Fix: Cable unplugged

Without dashboard: Spend 15 minutes SSH'ing and grepping logs

With dashboard: Find issue in 30 seconds

Performance Tuning

Robot feels sluggish:

Dashboard shows:

• path_planner using 85% CPU
• Tick rate dropped to 25 Hz (should be 60)
• Logs show "Path computation taking 40ms"

Action:

• Optimize path planning algorithm
• Or reduce planning frequency
• Or move to faster computer

You found the bottleneck in 2 minutes

Team Collaboration

Scenario: Two people debugging

Person 1 (at robot):

• Running the code
• Making physical adjustments

Person 2 (at desk):

• Watching dashboard
• "CPU just spiked to 95%"
• "Tick rate dropped when you turned that corner"
• "I see errors in the lidar node"

Real-time feedback loop without both people crowding around one screen

Access From Anywhere

The dashboard binds to all network interfaces, meaning:

Works great:

• Monitor from your desk while robot runs across the room
• Check on phone while robot is charging
• Team member in another room sees same view
• Demo to visitors on a tablet

How to access:

Same computer:

http://localhost:3000

Different device (same WiFi):

http://192.168.1.100:3000

(Use your computer's actual IP - dashboard shows it at startup)

Security note: No authentication by default. Only use on trusted networks.

Tips

Leave it open: Keep dashboard open while developing. Glance at it occasionally to catch issues early.

Second monitor: Perfect use for that extra screen - full-time system monitoring.

Mobile monitoring: Put it on a tablet mounted near your robot for at-a-glance status.

Parameter experiments: Export parameters before experimenting, so you can restore if needed.

Quick Reference

What each tab shows:

Health colors:

• 🟢 Healthy - Everything normal
• 🟡 Warning - Minor issues
• 🟠 Degraded - Some errors
• 🔴 Critical - Major problems
• ⚪ Idle - Nothing running

Troubleshooting

Dashboard shows nothing Make sure your HORUS app is running first

Can't access from phone Check both devices on same WiFi Allow port 3000 in firewall: sudo ufw allow 3000

Port already in use Use different port: horus dashboard 8080

Parameters not saving Create config directory: mkdir -p ~/.horus/config

Terminal UI Mode

Prefer the command line? Use terminal UI mode:

horus dashboard --tui

All the same features, rendered in your terminal. Perfect for:

• SSH sessions
• Headless servers
• Terminal enthusiasts

Next Steps

• CLI Reference - Dashboard commands
• Parameters Guide - Deep dive into params
• Examples - Real usage patterns