01
The time is unknown
The incident surfaces in a count, not from a witness. You are left with a wide time window.
Know what happened.Not just that it was recorded.
Your cameras already capture everything. ACALAS turns that footage into searchable incident intelligence — what happened, when, and what it cost. Entirely on your own infrastructure.
An illustrative example query. This field is not interactive.
SCROLL
THE PROBLEM
Searching costs more than the loss itself
The count comes up short. The weighbridge shows a gap. Damage turns up in the morning. Nobody knows when it happened. Scanning three weeks of footage across dozens of cameras takes days — and the loss rarely justifies it. So nobody looks. It gets written off. Everyone moves on.
What goes uninvestigated repeats. By year end the write-off line has grown and nobody knows where it came from.
02
Scanning is expensive
Reviewing footage takes days of someone's time. So most incidents are never looked at.
03
Footage is dead capital
The money spent on cameras produces evidence, but not decisions.
THE PLATFORM
A camera is a sensor. We turn it into data.
Conventional systems store footage. ACALAS reads it: which object, in which zone, for how long, at what time. The output is not a pile of clips — it is a queryable record of what happened.
SEARCH
LIVEFind an incident you cannot timestamp by describing the event instead.
INCIDENT RECORD
LIVEEntry, dwell and exit as one continuous timeline instead of scattered clips.
SHIPPINGIN DEVELOPMENT
EVIDENCE PACKAGE
IN PROGRESSYou will be able to export an incident's clips, timestamps and camera IDs as a single file — for insurance, legal, and audit.
CROSS-CAMERA TRACKING
IN PROGRESSYou will be able to follow a vehicle or person across cameras as one continuous story.
MULTI-SITE COMMAND
ROADMAPYou will see what is happening at remote sites from a single operations view.
ROADMAP
Today we find incidents.Tomorrow the facility explains itself.
ACALAS is not a search box. The goal is a system that can account for what happened across an entire operation — and eventually anticipate it. We are building it in five stages, and we say plainly which stage we are on.
01
FIND
LIVELocate an incident when you don't know when it happened. Place, action and duration are searched together.
02
TRACE
IN PROGRESSYou will be able to follow a subject across cameras and reconstruct the full sequence of an incident.
03
ALERT
ROADMAPYou will be told when a situation you defined occurs, instead of searching for it afterwards.
04
UNDERSTAND
ROADMAPFootage will become operational data: dwell times, bottlenecks, repeat violations. Security data will become operations data.
05
SCALE
ROADMAPHundreds of cameras across multiple sites will run under one sovereign deployment.
Every stage is measured on real footage before it ships. We do not present anything as working that we have not measured.
WHERE IT WORKS
Wherever loss goes unnoticed
The pattern is always the same: many cameras, long retention, and incidents discovered by a count rather than a witness. Nobody knows when it happened, so nobody looks.
The scale changes. The problem does not.
PRIMARY FOCUS
01
MANUFACTURING
Scrap movement, shift changeover, unexplained line stoppages.
Usually asked by operations or plant management.
- a truck loading scrap after shift end
- someone entering the mill floor during a stoppage
- a forklift idle at the silo for over 20 minutes
01Scrap movement, shift changeover, unexplained line stoppages.
Usually asked by operations or plant management.
- a truck loading scrap after shift end
- someone entering the mill floor during a stoppage
- a forklift idle at the silo for over 20 minutes
02Inventory gaps, dock dwell time, loading outside schedule.
Usually asked by warehouse management or finance after a stock count.
- a van at the loading bay outside delivery hours
- a person carrying boxes toward the staff exit
- a trailer waiting at the dock for more than 40 minutes
03After-hours access, material removal, delivery verification.
Usually asked by the site manager or after an insurance claim.
- a vehicle entering the site at night
- someone loading rebar onto a pickup
- how many mixers arrived on Tuesday
04Damage disputes, parts movement, unauthorised test drives.
Usually asked by the service manager during a customer dispute.
- a vehicle leaving the yard after closing
- someone opening the parts store door
- a car returning more than an hour after leaving
05Stockroom exits, staff exits, return disputes.
Usually asked by loss prevention or regional management.
- a person leaving through the stockroom door
- staff exiting with a bag after closing
- someone at the returns counter without a customer
The same problem exists at a different scale.
These environments share the same problem at a larger scale. The platform is being built toward them; we are not deployed there yet.
EXPANDING TO
PORTS & TERMINALS
Container movement, gate transits, cargo damage disputes
a container moved outside its scheduled window
CUSTOMS & BORDER CROSSINGS
Vehicle flow, inspection lanes, bonded warehouse access
a vehicle bypassing the inspection lane
AIRPORTS & AIR CARGO
Apron activity, baggage areas, perimeter integrity
a vehicle on the apron outside its assigned slot
ENERGY & MINING SITES
Remote unmanned areas, fuel movement, equipment accountability
fuel being drawn from a machine at night
CAMPUSES & MUNICIPAL
Large perimeters, distributed cameras, incident review
someone entering a restricted zone after hours
DATA SOVEREIGNTY
Your footage never leaves your site.
ACALAS runs on your own servers. No cloud, no external transfer, no dependency on an internet link. In camera systems the highest regulatory risk is transfer — and that risk closes at the architectural level, not with a policy document.
This was not added later. It is the founding constraint of the product. For ports, customs, energy and defence-adjacent operators — and increasingly for industrial groups under KVKK and GDPR — sovereignty is no longer a preference.
- ON-PREMISE
- The server is yours. The data is yours.
- NO NEW DATA CATEGORY
- Your retention and deletion policy is unchanged.
- NO BIOMETRICS
- Face recognition and person re-identification are deliberately not performed.
OUR LIMITS
We will also tell you what we don't do
Vendors rarely say where their system stops. We say it up front.
No face recognition
No person re-identification — it counts as biometric data
No counting
No licence-plate reading
No small-object detection at distance
What we do is place, action and time. We do that well.
Let's measure it on your own footage.
Before we sell you anything, we want to test it on your site. Give us a sample from an incident you already know about, and we will share the result in writing — whether it works or not.