NEW Feature
SNIPR Insights
Thank you for selecting OneSpace Central to empower your community. We at OneSpace are more than happy to assist you in case you have any troubles with your Site Management system.
This guide was made to provide a quick reference for common actions, functions, and workings within the Insights pages of Central. If this guide fails in assisting you with understanding the insights functionality, please do not hesitate to email us on support@one-space.co.za.
This guide will address the following:
The SNIPR Insights page gives you a real-time view of every vehicle and face your camera network has picked up — what was read, where, and whether it warrants attention. It combines a live camera activity map, an ANPR/Face event feed, a plate-origin breakdown by province, transit trend charts, and a predictive threat timeline that forecasts risk based on time of day, moon phase, and weather. Use the time range filter at the top to view activity over the last 24 hours, 7 days, 30 days, or 90 days.
Time Range & Key Stats
This bar sets the context for everything else on the page.
Time Range Filter: Choose the period you want to analyze: 24 hours, 7 days, 30 days, or 90 days. Every map, chart, and list on the page updates instantly to match your selection.
Key Stats: A quick-glance summary of camera network activity for the selected period: Reads (total plate/face captures), Unique (distinct vehicles seen), New (first-time plates), No-plate (vehicles detected without a readable plate), Daily Avg, Peak Hour, Alerts (watchlist/flag matches), Faces (facial recognition events), and Offline (cameras that dropped out during the period). Most cards show a percentage change against the previous equivalent period.
How to read it: Green ↑ means the number has grown since the last comparable period; red ↓ means it's dropped — note that for Alerts, a red ↓ is good news (fewer alerts), the opposite of how you'd read it for Reads or Unique. Peak Hour and Offline don't carry a comparison arrow, just a value.
What to do with it: Start every review here before drilling into the sections below. If Alerts or Offline is elevated, treat it as a trigger to jump straight to the ANPR Events feed or check camera health before anything else. A high No-plate count relative to Reads is worth a glance too — it can point to a camera angle, lighting, or plate-cover issue rather than genuine unreadable traffic.
Camera Activity Map
A heatmap of where your cameras are picking up activity, overlaid on the property boundary. Each pin marks a camera location; the warm red/orange glow shows where hits are concentrated, while cooler areas represent lighter, baseline activity.
View toggle: Switch the map between Reads (raw plate/face capture volume), Alerts (only flagged/watchlist matches), and Alert Rate (alerts as a proportion of reads, useful for spotting a camera with a disproportionately high hit rate rather than just high traffic).
How to read it: The brighter and redder a cluster, the more activity that camera has logged relative to the rest of the network. Switching to Alerts strips away routine traffic and shows you only where flagged events are actually happening — which isn't always the busiest camera.
What to do with it: Start on Reads to see where traffic is heaviest, then flip to Alerts to check whether that's also where your risk is concentrated — the two don't always line up, and a quiet camera with a spike in Alerts is often more worth investigating than a busy one with none. Use the enlarge icon (top right of the panel) to pull up a bigger view when you need to inspect camera positions or the boundary overlay more closely.
ANPR Events (Part 1: The Feed)
A live, feed of every plate read and face recognition event captured in the selected time range, newest first. Each entry shows the time, plate number (or No plate if unreadable), the camera/device that captured it, and a type tag (ANPR or Face).
How to read it: Use the Alerts and Face toggle buttons to narrow the feed to just flagged matches or facial recognition events, or use the Filter dropdown for more specific criteria. The thumbnail on the left is the actual captured image for that read.
What to do with it: Scan this feed at the start of a shift to review recent activity, or filter to Alerts to confirm nothing flagged needs immediate follow-up. Click on any row to drill into the full history for that plate — which is exactly what Part 2 below covers.
ANPR Event Details (Part 2: The Drill-Down)
Clicking any row in ANPR Events opens the full profile for that plate: an Overview (classification/province, total SNIPR reads, first seen, last seen), a gallery of ANPR Read Images (recent daytime captures), a Recent Site Reads timeline (every camera/time this plate was seen on site), and Gatebook Data — whether this plate matches a known vehicle registered to the estate, and how many Gatebook events exist for it.
Why this matters: A single row in the feed only tells you one moment. This view tells you the plate's whole story on your SNIPR Site — how often it's really been here, whether it's a known/registered vehicle, and whether it's ever matched a Gatebook entry. That's the difference between "a car drove past" and "this car has been here 7 times over 90 days and isn't registered to anyone."
What to do with it: Whenever a plate in the feed looks unfamiliar, appears repeatedly, or triggers an alert, drill in before deciding it's nothing. Pay particular attention to Gatebook Match: No match paired with a high SNIPR Reads count — a plate with no Gatebook record but frequent, recurring visits is a strong candidate for follow-up with security or the estate office.
Plate Type Breakdown
A donut chart showing where the plates your cameras have read are registered, broken down by province (with the total tagged reads shown in the center). Expand SA Province to see the split across all nine provinces, plus an Unknown category for reads that couldn't be classified.
Click to filter: Selecting a province in this list filters the ANPR Events feed to show only plates registered to that province — a quick way to isolate, say, every out-of-province plate seen today without building a filter manually.
How to read it: Each row shows a province's share of total tagged reads, the count, and its change versus the previous equivalent period. Since this is a residential estate, you'd expect the local province (here, KwaZulu-Natal) to dominate — the further down the list you go, the more those plates stand out as coming from further afield.
What to do with it: Use this to sanity-check the makeup of your traffic — a sudden jump in an out-of-province or Unknown share is worth a look, especially alongside a rise in Alerts on the Key Stats bar. It's also a useful long-range view (7/30/90 days): a growing proportion of unfamiliar-province plates over time can flag a broader pattern that a single day's reads wouldn't show.
Transit Trend
A combo chart plotting reads over time (bars) against the typical pattern for that period (dotted line), so you can see not just volume, but whether it's normal for the time of day.
How to read it: Each bar is the number of reads in that time slot; the dotted Historical Average line is what's typically expected at that same time. Bars rising above the line mean busier than usual; bars sitting below mean quieter than usual. The shape also reflects your site's natural rhythm — a dip overnight, rising again through the day.
What to do with it: Use this to plan camera monitoring and patrols around real peak times rather than guesswork — it's the detailed, hour-by-hour version of the Daily Avg and Peak Hour stats up top. More importantly, watch for bars that break away from the dotted line: a spike at an hour that's normally quiet (like the middle of the night) is exactly the kind of anomaly worth cross-checking against ANPR Events for that specific window.
Predictive Threat Timeline
A forecast chart that scores overall risk (0–100) for your site, combining the last 24 hours (solid line) with a 1-day-ahead forecast (dotted line), split by a now marker. The colored bands in the background — green/blue (low), yellow (elevated), red (high) — show which risk tier the score falls into at any point, and the Low Risk badge (top right) reflects where the score sits right now.
Signal Breakdown — Hover over any point on the timeline to see what's driving the score at that moment: Volume (how much activity), Time (how unusual the hour is), Moon (moon phase), and Weather — each shown as its own contributing value.
How to read it: A rising line moving into the yellow or red band means the model expects elevated risk soon, based on patterns like unusual timing, past incident correlation, or forecast conditions — not a single visible event. The Signal Breakdown tells you why the score is what it is at a given hour, rather than leaving it a black box.
What to do with it: Use the overall badge as your at-a-glance check — if it ever moves out of Low Risk, look at where the line is heading (rising into the forecast half) and check the Signal Breakdown to see which factor is driving it. If Time or Volume is the dominant signal heading into an odd hour, that's your cue to increase monitoring or patrols during that window before it happens, rather than reacting after the fact.