Your control room watches what the platform tells it to.
Not the other way around. A bespoke offsite monitoring platform — AI object and scene detection, SOP-driven response orchestration, integrated tour and incident management. Decisions, not guesswork.
For security companies, estates, and operations tired of relying on operators to spot what 32 screens have hidden — and ready to give them a platform that does the watching, so they can do the deciding.
Six things wrong with the way most control rooms run today.
The reactive monitoring model has been the standard for two decades. It worked when sites were small, cameras were few, and threats were predictable. Today's reality — dozens of cameras per site, multi-site responsibility, 24/7 coverage, audit-driven compliance — has outgrown the model. The cracks are showing in every control room that hasn't moved on.
Humans Can't Watch 32 Screens
Research is clear: after 20 minutes, an operator's effective visual attention drops by 95%. The math doesn't lie. Your operator is missing events — not because they're bad at the job, but because the job is biologically impossible.
Training That Walks Out
Months invested training a controller in your SOPs, your sites, your client expectations. They resign. The replacement starts from zero. The only system that holds the knowledge is the one that just left. Repeat every 12-18 months.
No SOP Enforcement
The SOP exists in a folder. The controller's interpretation of the SOP at 3am is whatever they remember. Every escalation is a different decision from a different person on a different shift. SOPs without enforcement are SOPs in name only.
Tour Management on Paper
Tour sheets ticked off, signed at end of shift, filed somewhere. Did the tour actually happen? Did every checkpoint get visited? At what time? In what order? Who knows. The audit trail is whatever the tour sheet says.
Incident Reports, 4 Hours Later
The incident happens at 02:47. The report is written at 06:30 when the supervisor arrives. Half the detail is forgotten. The video clip nobody pulled at the time is buried in 12 hours of footage. Late reports are weak reports.
OMS Standard, Informally
The South African OMS standard for offsite monitoring exists. Most operations claim alignment, few can prove it. The audit asks for documented SOPs, response time evidence, controller training records, escalation logs. The answer is usually "we have it somewhere."
Three ways to run a control room. Only one of them scales.
Every control room operator in South Africa is doing one of these three things. The one in the middle is what most operations have moved to. The one on the right is the future — and it's the only model that actually solves the structural problems with reactive monitoring.
Operator-Driven Monitoring
Operator stares at screens. Operator decides what's important. Operator remembers the SOP. Operator types the report. The platform is a viewer. The human is the entire system. Outcomes vary by shift, by operator, by mood.
Motion-Triggered Monitoring
Motion detection raises an alert. Operator looks at the alert. Operator decides if it's real. Operator follows whatever SOP they remember. Better than nothing — but every cat, leaf, and shadow generates an alert. Operator alarm fatigue sets in within weeks.
System-Orchestrated Response
AI detects the right things — persons in restricted zones, vehicles where they shouldn't be, perimeter breaches, abandoned objects. The platform classifies the event, triggers the SOP, presents the operator with the decision — not the search.
Operator as Decision-Maker
The platform does detection. The operator does decision-making. Right work for each role — not asking the human to be the detection layer.
SOP Encoded, Not Remembered
The SOP lives in the platform. Triggered by the event. Presented to the operator at the moment of decision. Consistent response, every time — regardless of operator, shift, or fatigue.
Lighter Training, Better Outcomes
The platform carries the cognitive load. Operators learn the platform, not 200 site-specific SOPs. New operators are productive in days, not months. Knowledge stays in the system, not in the resignation letter.
Eyes that don't blink. Across every camera. All the time.
Modern AI doesn't see "motion". It sees persons, vehicles, objects, behaviours, anomalies — and it sees them across hundreds of camera streams simultaneously, without fatigue, without bias, without the 20-minute attention drop that limits human operators. The platform watches. The operator decides.
Person Detection
Persons in restricted zones, after-hours presence, loitering above threshold time, persons in groups vs alone. Detection plus context.
Vehicle Detection
Vehicles in no-stop areas, vehicles after curfew, unknown vehicles vs registered, vehicle direction-of-travel against expected flow.
Perimeter Breach
Line-cross detection at fence boundaries, intrusion zones, virtual tripwires across approach paths. Pre-perimeter detection where camera coverage allows.
Abandoned Object
Bags, packages, vehicles left in monitored areas above a configurable dwell time. The detection most reactive systems miss completely.
Climbing & Scaling
Persons climbing fences, walls, gates, structures. Detection happens during the act — not after the breach is complete.
Loitering & Behaviour
Persons remaining in monitored areas above expected dwell time, repeated approaches, suspicious movement patterns flagged for operator review.
Crowd & Group Detection
Group formation in restricted areas, crowd density above thresholds in public spaces, gathering pattern detection at sensitive locations.
Object Removal
Removal of monitored items — equipment, vehicles, infrastructure — from defined zones. Theft detection at the moment of action.
Custom Scene Rules
Site-specific rules combining multiple conditions — "person + after-hours + in zone X + dwell > 30s" — built per site, tuned to your reality.
The platform tells the operator what to do next.
An event happens. Without a platform, the operator has to remember the SOP, decide the right action, choose the right escalation, type the right report. The platform handles all four — presenting the operator with the next required action, the right contact, the script to follow, the boxes to tick.
Detect
AI flags an event. Classified, prioritised, geo-tagged.
Route
Platform applies the SOP for this event, this site, this time.
Present
Operator sees the event, the SOP, the next required action.
Site, Time, Type, Severity
Same event type can route differently depending on site, time-of-day, day-of-week, severity classification. A person in zone A at 10:00 routes to a different SOP than the same person at 22:00.
Step-by-Step Operator Guidance
Each SOP step presented in sequence. Tick-box completion. Mandatory fields enforced. Script wording for client calls. Photo/note evidence capture inline. The operator follows the SOP because the platform won't proceed without it.
Right Contact, First Time
The platform knows who to call — site contact, client supervisor, response team, control room manager. Escalation path defined per site, per event type, per time-of-day. No more "let me find the contact list."
Every Action, Time-Stamped
Every operator action logged. Every escalation recorded. Every photo and note attached. The audit trail builds itself as the SOP runs — no end-of-shift report writing, no missing detail.
SOPs in folders are SOPs in name only. SOPs encoded in the platform are SOPs that actually run — every event, every operator, every time. Consistency becomes infrastructure, not a hope.
— The Swiss Systems SOP positionVirtual tours. Physical tours. Both verifiable.
Tours are the most-claimed and least-verified part of any monitoring operation. Did the tour happen? Did every checkpoint get visited? At what time? The platform answers all three — for virtual tours run from the control room, and for physical tours run by guards on the ground. The tour sheet is replaced by an audit trail that builds itself.
Camera-Based Site Walks
Scheduled virtual tours run from the control room. Operator steps through a defined sequence of cameras, observes per checkpoint, ticks off completion. Time-stamped, audit-trailed, auto-escalated if not completed.
- Per-site tour schedules
- Per-checkpoint observation requirements
- Photo capture at key checkpoints
- Notes per checkpoint
- Auto-escalation on missed tour
Ground-Truth Verification
Mobile app for guards on the ground. NFC tag, QR code, or GPS check-in at each checkpoint. Time-stamped, photo-evidenced, signed. Real proof the tour actually ran — not a tick on a paper sheet at end of shift.
- NFC / QR / GPS checkpoint verification
- Mandatory photo capture
- Off-route detection
- Missed-checkpoint alerts
- Live tour visibility from control room
Configurable, Auditable
Tour schedules built per site — frequency, timing, sequence, randomisation. Audit-friendly: client can ask "show me every tour at site X over the last 30 days" and get a documented answer in seconds.
- Frequency: every X minutes, hours, shifts
- Sequence: fixed or randomised
- Time-conditional schedules
- Site-specific patterns
Missed Tour, Late Tour, Off-Route
Missed checkpoint? Auto-escalate. Late check-in? Auto-flag. Guard off-route? Visible to control room. The exceptions get caught at the moment they happen, not at end-of-shift reconciliation.
- Real-time exception alerts
- Configurable tolerance windows
- Auto-escalation paths
- Supervisor visibility
Capture at the moment. Not at end of shift.
The single biggest failure of incident management in reactive control rooms is the gap between event and report. By the time the report is written, half the detail is forgotten and the video evidence is buried. The platform closes the gap — incident capture happens at the moment of detection, with structured intake, evidence pre-attached, escalation pre-routed.
The Right Questions, In Order
Every incident type has a defined intake form — what happened, where, when, who responded, what was the outcome. Mandatory fields enforced. Free-text fields kept to a minimum — structured data is searchable, free text isn't.
- Per-incident-type intake forms
- Mandatory field enforcement
- Drop-down classifications
- Auto-populated time, location, operator
Video Clips Auto-Bookmarked
The video clip from the moment of the event auto-bookmarked into the incident record. Photos captured during SOP execution attached. Notes timestamped. The evidence pack assembles itself — no manual pulling, no buried clips, no missing context.
- Auto video clip bookmarking
- Inline photo capture
- Time-stamped operator notes
- Evidence pack PDF export
Right Person, First Time
Escalation contacts defined per incident type, per site, per time. Platform routes the escalation automatically, with the incident summary, the evidence link, the SOP step the operator is on. One click from detection to escalation.
- Per-site escalation paths
- Time-conditional routing
- SMS / email / call dispatch
- Acknowledgement tracking
Beyond the Single Incident
Every incident becomes a data point. Most common incident types per site. Peak periods. Repeat patterns. The system that catches one event also catches the trend across hundreds. Monthly reporting, client-facing dashboards, audit-ready exports.
- Per-site incident analytics
- Trend detection across periods
- Client-facing dashboards
- Audit-ready PDF exports
A layer on top — not a rip-and-replace.
You've already invested in cameras, NVRs, and a VMS. The Swiss Systems Offsite Monitoring Platform is designed to sit on top of what you already have — pulling video streams from your existing VMS, applying AI detection, orchestrating SOPs, building the audit trail. No camera changes. No VMS migration. No infrastructure write-off.
The fastest way to fail a control room migration is to insist on replacing everything. The fastest way to succeed is to layer the new capability on top of what already works — and let the rip-and-replace conversations happen later, on their own merits.
— The Swiss Systems integration positionVMS-Agnostic Stream Pull
Direct stream pull from any major VMS or NVR. RTSP, ONVIF, vendor-specific APIs all supported. Live streams flow into the AI detection layer without the VMS even noticing.
AI Detection Layer
Detection runs in our infrastructure, not on your cameras or NVRs. No edge AI requirement, no camera replacements, no on-prem GPU spend. Your existing camera estate becomes AI-capable overnight.
SOP & Escalation Layer
Detected events flow into the SOP orchestration layer. Operator sees both the live video (from your existing VMS) and the platform UI (from us) side by side.
No Infrastructure Write-Off
The cameras, NVRs, recorders, network equipment you've already paid for — all preserved. Your client base, your existing service contracts, your team's training — all preserved. Augmentation, not displacement.
Built to the standard. Designed beyond it.
The South African OMS standard for offsite monitoring sets the floor. Most operations claim alignment, few can prove it under audit. Swiss Systems is built to the standard from the ground up — and engineered to exceed it in three specific areas where the standard is now showing its age.
What the OMS Baseline Requires
- Documented SOPs per site and incident type
- Trained controllers on each operational site
- Documented escalation procedures
- Tour management with checkpoint verification
- Incident logging with time-stamps
- Evidence retention per regulatory requirement
- Audit trail for events and operator actions
- Client reporting at agreed intervals
Where Swiss Systems Exceeds
- AI-driven detection beyond motion-based alerting
- SOP enforcement at the moment of action, not at audit
- Real-time evidence packaging, not retrospective assembly
- Trend analytics across incidents, sites, operators
- Client-facing dashboards beyond monthly PDFs
- Single audit trail covering tours, incidents, escalations
- Per-operator performance visibility
- Continuous improvement loop via incident trend data
Standards exist because the industry needed a floor. The floor is not the ceiling. The platforms that win the next decade of contracts will be the ones that take the standard as the starting point — and engineer well past it.
— The Swiss Systems standard positionThe platform that runs on infrastructure we own.
Most monitoring platforms run on third-party clouds — AWS, Azure, somebody offshore. When something breaks at 3am, the support phone tree starts. Swiss OMS runs in our own South African data centre, on our own infrastructure, supported by our own engineers. One throat to choke. Local data residency. Real accountability.
We Built the Platform
Bespoke software, written in-house by Swiss Systems engineers. No white-label, no licensed-from-overseas, no "we resell their product". The roadmap, the bug fixes, the feature requests — all under one roof.
We Host the Platform
Same data centre that hosts 20,000+ cameras of customer VMS workloads. Same redundant power, same redundant fibre, same engineered cooling. Production-tested infrastructure — not a sales claim.
We Run the Platform
Same engineers who built it monitor it 24/7. When you raise a ticket, you don't queue with general support — you reach the team that knows the platform end-to-end. No vendor handover, no offshore escalation.
Less training. Better outcomes. Higher retention.
The structural shift from operator-driven to platform-driven monitoring isn't a feature change — it's an operating model change. The role of the controller changes. The role of the supervisor changes. The economics of the entire control room change.
Decision-Maker, Not Watcher
Stops being the detection layer. Starts being the response layer. Cognitive load drops, decision quality rises. New controllers productive in days, not months. Job satisfaction goes up because the work is meaningful — not just staring at screens.
Coach, Not Catcher
Stops being the SOP enforcer. Starts being the trend-spotter. The platform catches the operator who skipped a step — automatically. The supervisor's time is freed for coaching, client engagement, and continuous improvement.
Strategist, Not Firefighter
Stops worrying about whether SOPs are being followed. Starts worrying about whether SOPs are the right ones. Real visibility into operations, real data for client conversations, real evidence in tender responses.
The control room of 2010 was a room full of screens and a person trying to watch them all. The control room of 2025 is a platform doing the watching, with people doing the deciding. Same physical space — completely different operating model.
— The Swiss Systems operating model positionWe built it. We host it. We support it.
The offsite monitoring market in South Africa has a dozen platforms — most of them resold from overseas, most of them generic, most of them assuming you'll do the integration work yourself. We built our platform from the ground up, in South Africa, for the South African operating reality. Same team builds, hosts, supports — and accepts single accountability for the whole thing.
We Built the Platform
Not a white-label. Not a reseller licence. Bespoke software, written by Swiss Systems engineers, owned by Swiss Systems. Roadmap, bug fixes, feature requests — all under one roof. Our flagship — and we treat it like one.
Built for SA, Not Adapted
Designed around the South African operating reality — load-shedding, multi-site security companies, the OMS standard, local data residency requirements, local response logistics. The platform doesn't need to be adapted — it was built here, for here.
Layer, Not Rip-And-Replace
Works with your existing cameras, NVRs, and VMS. No infrastructure write-off. No camera changes. No VMS migration. Augmentation, not displacement — the fastest path to real value from a deployment.
AI & SOP Together
AI without SOP orchestration is just better alerts. SOP without AI is paper in disguise. The combination — detection plus enforcement — is what changes the operating model. Most platforms have one. We have both, integrated.
Hosted In Our Data Centre
99.9% uptime SLA. 100 kVA redundant power. 10 Gbps redundant fibre. Same data centre running 20,000+ cameras of customer VMS workload. Production-tested infrastructure — not a sales claim.
Whole-Stack Partner
Offsite monitoring sits alongside our data centre, private cloud, VOIP, networking, M365, custom development, and managed IT services. One accountable partner for everything technology-related — no juggling four vendors.
Book a free control room assessment.
Half a day with a senior solutions engineer. We'll walk your control room, audit your current operating model, identify the structural gaps, and hand you a written assessment — including specific recommendations and a deployment outline. Yours to keep, regardless of whether we end up working together.
Walk your operation
A real conversation about how your control room runs today — operators, SOPs, tours, incidents, audit position. No assumptions, no upselling.
Find the structural gaps
Where reactive monitoring is costing you. Where SOPs aren't being enforced. Where audit exposure is sitting. Documented, prioritised, yours to keep.
Sketch the deployment
If we work together, what would the rollout look like? Timeline, integration points, training needs, expected outcomes. Realistic, not rose-tinted.