About TrailGuard and Nightjar IQ.

A conventional camera trap stores images for you to collect later. TrailGuard decides in the field. AI on the camera judges whether what triggered it matters, and if it does, the image reaches your team in about 30 seconds. Most teams run both, and use TrailGuard where response time changes the outcome.

You stop reviewing empty frames, because the filtering happens on the camera before anything is sent. You find out in time to answer a fence breach, not after it. You can also see which cameras are low on power and which have stopped reporting.

Conventional traps stay useful for density and survey work, where you will collect cards on a schedule anyway. Nightjar IQ manages TrailGuard units and not third-party cameras, so a mixed fleet means two systems.

Every transmitted image is classified and kept in your record on Nightjar IQ, not only the ones that match an alert rule. The same deployment that runs your anti-poaching alerts builds a continuous species and site record you can query, export, and report from later.

Teams reporting under a biodiversity obligation draw on that record, not on the alert history.

Most platforms are retrospective. You upload imagery and get analysis back, which is no help on the day someone cuts your fence. The second difference is ownership. We take no perpetual licence over your imagery, we don’t sub-license it to third parties, and we delete your raw data 30 days after you leave.

Platform terms in this sector vary more than they appear to, which matters on wildlife crime, sensitive species locations, and land under commercial agreement. Ours are in our standard sales agreement, not in a click-through.

If retrospective biodiversity analysis is all you need, there are good tools for it and we’ll say so. TrailGuard is built for teams who have to respond, and who need the record afterwards to hold up.

Yes. That is what Nightjar IQ is for. The same detections serve several jobs at once, and roles control who sees what. Response teams get only the alerts that need action, often in a Telegram group on a phone, with no dashboard involved.

Whoever runs operations sees fleet health: which cameras are quiet and which are low on power. Ecologists get the full species record, including everything that never triggered an alert. Reporting leads pull structured exports when a funder or regulator asks. Several sites sit under one account, so a programme running multiple landscapes manages them together.

Where a team already works in another tool, detections push there instead, into EarthRanger or into your own systems by webhook.

About 30 seconds from the trigger where cellular coverage is good. That covers capture, on-camera filtering, transmission, classification and delivery to Telegram or email. Weak signal extends it. A camera with no signal will not alert at all, and you collect the card in the field.

Units ship in 1 to 3 weeks. While they are in transit we set up your project, so the platform is ready when the hardware lands, and most deployments are running within days of it arriving. Each camera takes a few minutes to mount, frame, and signal-check.

After that you are configuring, not installing: which detection classes transmit, which rules alert whom, whether detections push into EarthRanger or your own systems. We run an onboarding call with your team and stay on a direct channel through the first weeks in the field. Questions such as why one camera is quiet and another noisy only surface once units are out there.

TrailGuard needs a cellular signal. Each camera takes a standard local SIM and works on mobile networks worldwide. You supply the SIM and the airtime. A 1GB monthly plan is usually enough for a camera in normal operation, and we can advise on carrier choice for your region.

Yes. TrailGuard is carrier-agnostic and takes any standard local SIM, with APN settings configurable during setup. Use whichever network performs best at your site, or an account you already hold. What we can’t do is warrant coverage or carrier performance at any particular location.

So test the signal at the mounting point itself, not on a coverage map. Where two carriers overlap, trial both on a single camera before you commit a fleet.

TrailGuard needs cellular today. Non-cellular connectivity is the capability customers ask for most and the most active area of our development work. Field trials are running now and we expect wider availability in the coming months. You can’t buy it off the shelf yet.

If you are planning a deployment in a no-coverage area, talk to us early and tell us where. In some cases we can bring you into the trial.

Weak signal is workable. Transmissions retry, which drains batteries faster, and in marginal conditions some alerts won’t get through. TrailGuard has a signal test in setup mode, so you can check a mounting point before you commit to it.

Test the position itself, move the camera if the reading is poor, and consider a higher-gain antenna where the terrain fixes where the camera has to go. We’ll walk your team through the thresholds during onboarding.

Where there is no signal at all, the camera keeps triggering, filtering, and storing to its card. You collect the imagery in the field instead of receiving alerts.

It depends on what you are detecting. Wildlife is usually picked up within about 12 metres. People are larger and warmer, so they are detected further out. Vehicles trigger further again, because a warm engine is a much stronger heat signal than a walking animal.

Mounting height and downward angle set the far limit, because the sensor’s field eventually meets the ground. We’ll advise on mounting for your use case, and the setup preview lets you confirm framing on site before you lock the bracket.

Point TrailGuard along the trail, not across it. The motion sensor sees a narrower cone than the camera lens, so subjects need to pass through the centre of the frame and not clip the edge. Framing is the most common cause of a deployment underperforming.

In practice:

• Mount 1.5 to 3 metres up, angled slightly down, with more ground than sky in the frame
• Sit 7 to 10 metres back from where subjects will pass, so they cross the centre of the detection zone and not its far edge
• Face north where you can, or at least away from sunrise and sunset, to cut glare and heat-driven false triggers
• Mount vertically with the cable running upward, which is easier to weatherproof and conceal
• Use the live preview to frame the shot before you lock the bracket

TrailGuard is built as two bodies, not one. Only the camera sensor needs to face the trail, and it’s small: a palm-sized head in a camouflaged pattern that mounts flush against bark. The communications unit, battery, and antennas live in a separate body you hide higher up the tree or behind it, connected by a thin cable that runs like a vine up the trunk. The visible footprint at the target’s eye level is a fraction of a conventional trail camera.

The infrared illumination is covert. Nothing flashes when it fires at night, which is one of the most common reasons trail cameras get found and removed.

For higher-risk sites we are developing lockable anti-tamper mounting. Cameras keep working when partly obscured. The sensor and lens need a clear line to the trail, but everything else can be well hidden.

If a camera is removed, you will know that day. It stops reporting its daily health check, and the dashboard shows the gap without waiting for your next site visit.

It depends on the routes, not the area. TrailGuard is built for choke points: trails, tracks, gates, fence lines, and water points. It is not built for blanket coverage of open ground. A small number on the routes people use will outperform a larger number scattered broadly.

TrailGuard is not a perimeter surveillance system. Tell us the size and shape of your site and where incursions happen, and we’ll advise on a realistic deployment instead of quoting a number.

About 10,000 transmissions per charge, which for most deployments means many months in the field. Battery life is driven by how much the camera transmits, not how much it detects. A quiet site with strong signal runs far longer than a busy one, or than a site where weak signal forces the modem to retry.

If endurance is critical: stop transmitting detection classes you don’t need alerted, raise the confidence threshold, and place cameras where signal is strong.

Yes, and they are two separate controls. What transmits is set on the camera: you choose which detection classes leave the field at all, such as animal, vehicle or human. What alerts is set on Nightjar IQ, where rules go down to species level and combine with conditions.

Anything you exclude from transmission is still captured and stored on the card. It doesn’t consume airtime or battery. On the platform side, one deployment can route an incursion alert to a response team and send wildlife detections only to the ecologists, on different channels and at different thresholds.

Most teams use both: transmit narrowly to protect the battery, then alert precisely on what comes through.

Almost never. Every TrailGuard camera reports a daily health check and a regular heartbeat, so battery level, signal strength and last-seen status for the whole fleet show on the dashboard without anyone leaving the office. If a camera stops reporting, you know that day.

Firmware updates are delivered over the air. Configuration is changed remotely and takes effect without a field trip, including transmission classes, thresholds, and alert rules. For teams running cameras across terrain that takes a day to reach, that removes most of the field visits.

No. TrailGuard captures stills only, a short burst of images per trigger. Video doesn’t fit the power and bandwidth budget of a covert, battery-powered device in a remote place, and it would break the economics of transmitting over cellular. There is no live streaming and no RTSP.

Occasionally, yes. The sensor responds to heat. Strong sun on bright ground such as snow, pale rock or bare sand, combined with moving shade, can briefly imitate a warm body crossing the frame.

Most never reach you, because the on-camera filter removes the large majority of captures before anything is transmitted. Where one camera is producing daytime noise, the cause is nearly always framing, and re-aiming it away from bright open ground fixes it. Our support team will look at the pattern with you.

Only people you authorise. Your detections sit in your own tenant on Nightjar IQ, isolated from other customers, and roles determine who sees what. Alerts go only to the recipients you configure. A small number of Nightjar staff can reach accounts for support, and that access is audited.

There is no shared view and no cross-customer visibility of any kind. If you add someone to a Telegram alert group, they see the alerts routed to that group and nothing else. An administrator on your account can revoke access at any time.

No. There is no public data layer in Nightjar IQ, and nothing is published or made discoverable by default. Your detections, camera coordinates and imagery stay inside your own tenant and reach only the people you authorise. TrailGuard embeds no GPS data in alerts or images.

A camera deployed covertly in a contested area should not carry its own coordinates in every image it sends. You enter coordinates in the dashboard instead, where they are used for mapping and integrations and never travel with the imagery.

Where people appear in frame, two further controls are available per project: automatic privacy blurring, and full suppression of human transmission, so only metadata leaves the camera and the image stays on the card.

TrailGuard data is stored on Google Cloud Platform, in a United States region by default. Other regions are available on request at no functional cost. We deploy to Frankfurt for customers with EU residency obligations, and to Asia-Pacific regions on the same basis.

Tell us the requirement before deployment and we’ll provision your project in the right region from the start.

You own your raw imagery, sensor outputs, and metadata throughout. If your agreement ends we keep that raw data for 30 days so you can export it, and after that it may be deleted. Exports run in open formats: CSV, Camtrap DP, and GBIF-compatible packages.

We retain rights to derived data, meaning the analytics and model outputs produced through the platform, and we use anonymised, aggregated versions to improve the product. That use never identifies you, your people or your deployment locations without written consent, and you can opt out of external use entirely.

We take no perpetual licence over your imagery and no right to sub-license it to third parties.

Not today. TrailGuard is cloud-managed and the classification pipeline depends on managed cloud services, so we don’t support a fully self-hosted deployment.

For large government and enterprise programmes with sovereignty requirements there is middle ground: regional hosting, and outbound forwarding of detections into systems you control. Tell us the specific requirement and we will tell you what is possible.

We can execute a Data Processing Agreement and provide our sub-processor list on request. EU-region hosting is available, and cross-border transfers run under standard contractual clauses at the infrastructure layer.

For imagery containing people, two controls ship today and are configurable per project: automatic privacy blurring, which obscures detected people before the image is stored or displayed, and full suppression of human transmission.

Yes. Our System Architecture and Data Security Overview is the first document most reviews ask for, and we’ll send it on request. We can also complete a CAIQ or SIG Lite questionnaire. Nightjar doesn’t hold its own SOC 2 or ISO 27001 certification. Our infrastructure runs on Google Cloud Platform, which is certified to both.

GCP holds ISO 27001, SOC 2 Type II, ISO 27017, and ISO 27018. Access control, tenant isolation and role-based permissions are documented in the overview.

Partly. Timestamps on live alerts are applied by our servers when the detection arrives, and they are reliable. Images recovered from the SD card carry the camera’s internal clock, which has to be set correctly at deployment. We supply a setup guide. An unset clock will date the imagery wrongly.

A dedicated evidence-export package with formal chain of custody is not built yet. It is on the roadmap and enforcement users are driving it. If you have specific admissibility requirements, tell us what they are. That input is shaping what we build.

1 to 3 weeks from order confirmation. TrailGuard ships EXW, and shipping is charged at cost through FedEx or DHL with no mark-up. Import duties, taxes and clearance are the buyer’s responsibility, and the cost varies a lot by destination, so ask for an estimate before you commit.

We hold some stock regionally, including in the United States, which can shorten lead time. Full customs and export documentation goes with every shipment.

We ship worldwide, and the buyer or their local partner handles import. In markets with device-certification regimes, that certification is part of the import process and sits with the importer, not with us.

If you are working through a specific regime, talk to us early. We have been through it with buyers in South Asia, the EU, Southern Africa, and Southeast Asia, and can usually point you at what has worked, including freight-forwarder routes other customers have used.

Yes. We work with government and institutional buyers on purchase-order terms, and can supply what your procurement process requires: technical datasheets, manufacturer authorisation letters, export and customs paperwork, and the security architecture overview for IT review.

Tell us what your tender or procurement portal needs and we’ll prepare it.

12 months from dispatch, covering defects in materials and workmanship. We pay the shipping cost of returning defective goods. Misuse, modification, environmental damage and use outside design parameters are excluded.

Before a unit comes back, contact us with the device identifier and what you are seeing. Many apparent hardware failures turn out to be configuration or signal issues we can fix remotely, often the same day.

CSV for general analysis, plus Camtrap DP and GBIF-compatible packages for ecological and biodiversity reporting, all from the dashboard. If you report under IFC PS6, TNFD or CSRD, the structured detection record with server-applied timestamps is built to be the evidence layer those frameworks ask for.

Yes. Detections flow into your own EarthRanger feed with species, confidence, image and a link back to the platform, so your team works in the interface it already uses. Your EarthRanger administrator creates the event type and connection, which produces a key you enter in your alert rules.

Cameras need their coordinates recorded first, so alerts map to the right place. Telegram and email alerting keep running alongside it.

Platform webhooks are coming soon on paid tiers. When they ship, detections will push in real time to any endpoint you control: ranger dispatch tools, incident systems, custom alerting. EarthRanger integration is already live if that route fits.

Telegram and email today, with rules set by species, detection class, confidence, and camera. Telegram group chats work the same way as individuals, so a whole response team shares one channel. We are working on WhatsApp. It is not available yet.