Skip to main content

Initial Setup

After installing Anava, this guide walks you through creating your first detection configuration. By the end, you'll have a working setup that analyzes camera events and can trigger alerts in your VMS.

Understanding Detections​

Detections are the core concept in Anava. A detection defines:

  • What triggers analysis - Motion, object detection, digital input, schedule, or manual trigger
  • What to look for - People, vehicles, packages, fire, weapons, or custom objects
  • What questions to answer - Count people, detect loitering, identify activities
  • When to run - Always, after hours, office hours, weekdays, or weekends
  • What actions to take - Alerts, voice announcements (talkdown), webhooks

Detection Architecture​

Anava organizes detections using three components that work together:

Group structure showing skills, profiles, and devices

ComponentPurposeExample
GroupContainer that holds configurations and devices"Warehouse North"
SkillDefines what AI looks for and how it responds"Weapon Detection"
ProfileLinks triggers (motion, I/O) to skills with schedules"Analyze on motion, weekdays 6pm-6am"
Key Insight

Devices inherit all profiles and skills from their group. Change the group configuration once, and all cameras in that group are updated automatically.

Visualizing Your Detections​

The Detections page in the sidebar provides a visual map of how these components connect. Use it to:

  • See all groups and their profiles at a glance
  • Understand which skills each profile uses
  • Navigate quickly by double-clicking any item to edit it

Step 1: Create a Group​

Groups are the foundation of your configuration:

  1. Navigate to Groups in the left sidebar
  2. Click + Create Group
  3. Enter a descriptive name (e.g., "Building A - Entrances")
  4. Add an optional description
  5. Click Create

Group Naming Best Practices​

Good NamesAvoid
"Warehouse Loading Docks""Group 1"
"Retail - Checkout Area""Cameras"
"Parking Lot - Employee""Test"

Step 2: Assign Cameras to the Group​

  1. Click on your newly created group
  2. Go to the Devices tab
  3. Click + Add Devices
  4. Select cameras from the list
  5. Click Assign

Cameras will begin receiving the group's configuration within 30 seconds.

Step 3: Create Your First Skill​

Skills define what the AI analyzes and how it responds.

  1. While in your group, click the Skills tab
  2. Click + Create Skill

Configure the Skill​

Basic Information:

  • Name: "Intruder Detection" (or your use case)
  • Description: Brief explanation of what this skill detects

Analysis Configuration:

The skill defines prompts that guide AI analysis:

FieldPurposeExample
Pre-filter CriteriaQuick check before full analysis"Is there a person visible?"
Full Analysis PromptDetailed analysis instructions"Analyze this scene for unauthorized persons..."

Objects to Detect:

Add objects that should trigger ONVIF events for your VMS:

  • Click + Add Object
  • Enter object names: "Person", "Weapon", "Vehicle"
  • Enable Stateful if your VMS needs is-true/is-false state

Questions (Optional):

Questions generate structured data outputs:

QuestionTypeExample Use
"Is this person authorized?"BooleanAccess control integration
"How many people are present?"IntegerOccupancy tracking
"What is the person wearing?"StringIdentification support

Save the Skill​

Click Create Skill to save your configuration.

Step 4: Create a Profile​

Profiles link triggers (camera events) to skills (AI analysis).

  1. Click the Profiles tab in your group
  2. Click + Create Profile

Configure the Profile​

Basic Settings:

  • Name: "Motion-Triggered Intruder Detection"
  • Skill: Select the skill you created
  • Active: Enabled

Trigger Configuration:

Trigger TypeUse Case
MotionGeneral motion detection
Object AnalyticsAXIS Object Analytics scenarios
Digital InputExternal sensors, door contacts
ManualVirtual input triggers from VMS
ScheduleTime-based polling (Learning Mode)

For a basic setup, select Motion trigger.

View Area:

Select which camera view area triggers analysis:

  • View Area 1 (default full view)
  • View Area 2-8 for multi-stream cameras

Schedule:

Set when the profile is active:

  • 24x7 - Always active
  • Select Schedule - Use camera-defined schedules

Save the Profile​

Click Create Profile to activate your detection.

Step 5: Verify It's Working​

Test the Detection​

  1. Create motion in front of a camera in your group
  2. Navigate to Live Events in the left sidebar
  3. You should see a new event appear within 5-10 seconds

Event Details​

Click on an event to see:

  • Frame Captured: Image analyzed by AI
  • Analysis Results: Objects detected, questions answered
  • ONVIF Events: Events emitted to your VMS

Check Camera Status​

In the group's Devices tab, verify each camera shows:

  • Status: Online
  • Last Event: Recent timestamp
  • Active Profiles: Your profile name

What's Next?​

Now that you have a basic detection working:

Integrate with Your VMS​

Connect Anava's ONVIF events to your video management system:

Configure Notifications​

Set up alerts for your detections:

Scale Your Deployment​

Manage larger camera fleets:

Common First-Time Issues​

No Sessions Appearing​

  1. Verify the profile is set to Active
  2. Check camera is online in the Devices tab
  3. Ensure trigger type matches your test (motion vs I/O)
  4. Check schedule allows current time

Analysis Results Empty​

  1. Verify skill has a valid analysis prompt
  2. Check that objects are configured
  3. Ensure camera view isn't obstructed

VMS Not Receiving Events​

  1. ONVIF events only emit when objects are detected
  2. Ensure objects are set to Stateful if VMS expects boolean events
  3. Check VMS is listening to the camera's ONVIF event stream

For more help, contact support at support@anava.ai.