Limits

Honest about what it cannot see.

Every system has limits. Vantage shows you its limits inside the reading itself, room by room, instead of hiding them behind a guess.

The one rule

When unsure, say so.

A false clearA crew walks past a room with someone in it.

The worst answer Vantage could give.

A no readThe room is searched by hand.

So Vantage is tuned to say no read long before it would risk a false clear.

Conditions

Where confidence drops, and what to do about it.

These are the conditions we see most. In each one, Vantage tells you in the reading itself.

01

Dense concrete, steel and deep debris

No read

Readings cannot form through heavy, dense construction.

What to doSearch those rooms by hand.

02

Many people in one room

Range

Individual counts blur together in a crowd.

What to doTreat the count as a range, not a total.

03

A still, silent person

Resolving

Breathing needs time and calm conditions to confirm.

What to doWait for the confirm, or treat the room as occupied.

04

Heavy movement nearby

Low confidence

Machinery, fans, moving water and other crews lower confidence.

What to doPause nearby movement when you can.

05

Animals

Occupied

Large animals can read as occupants.

What to doTreat every reading as a lead, not a conclusion.

06

A building it has not read before

Cautious

Every structure is different, so Vantage starts cautious in a new one.

What to doExpect more low and no reads early on.

07

Below grade and flooded spaces

No read

Basements and flooded rooms may not resolve.

What to doKeep them on the manual search plan.

08

Large, open buildings

Coverage

Coverage is room-sized, not campus-sized.

What to doRead the rooms that matter most first.

Tuned toward caution

Being wrong costs more one way than the other.

When the evidence is weak, Vantage moves a room toward low confidence or no read, never toward clear. We would rather send your crew into an empty room than past an occupied one.

  • Weak evidence lowers the state, it never raises it
  • Unknown rooms stay on the search plan
  • Thresholds are set per deployment, with your team

How we test

Every pilot ends with the misses on the table.

Results from a lab do not tell you how Vantage behaves in your buildings. So every evaluation runs where you work, against ground truth your team controls.

  1. In your buildings

    We read the structures your crews actually enter, not a showroom.

  2. Ground truth you control

    Your team places role players. The operator does not know where.

  3. Every miss reported

    How often Vantage was right, how often it said no read, and every time it was wrong.

  4. Your decision

    You decide whether it earns a place in your operations. Nothing is assumed.

Questions about limits

What skeptics ask. We welcome it.

Why not just show a best guess?

Because a guess that looks like a reading gets trusted like one. A no read keeps the room on your search list, where it belongs.

Can we see the numbers from our pilot?

Yes. Accuracy, no-read rates and every miss from your evaluation are shared with your agency in full.

Does Vantage ever get a room wrong?

Yes, as every system does. That is why each reading carries a confidence, why weak evidence pushes toward no read, and why crews still search and clear every room.

What does Vantage never do?

It never produces images, never records audio, never identifies anyone, and never opens a session without a recorded basis.

Confident · 81% Low · 52% No read · concrete Clear · 94%
Vantage · limits

Limits, in public.

Ask us where Vantage struggles. It is the first thing we cover in every briefing.