MAJETE WATCH Wildlife Detection & Survey System
MAJETEWATCH.COM

Wildlife Detection & Survey System

A real-time AI-powered monitoring platform protecting one of Africa's greatest conservation success stories.

Majete Wildlife Reserve

Majete Wildlife Reserve is one of Africa's most remarkable conservation success stories. Located in the Lower Shire Valley of south-western Malawi, the 700 km² reserve was first proclaimed a protected area in 1955. By the late 1990s, decades of poaching and neglect had left it nearly devoid of wildlife — elephants, lions, rhinos, and most large mammals had been eradicated entirely.

In 2003, African Parks entered into a public-private partnership with Malawi's Department of National Parks and Wildlife to rehabilitate and manage the reserve. What followed is a story of extraordinary restoration. More than 3,000 animals from 17 species have since been reintroduced, including black rhino (2003), elephant (2006), lion (2012), giraffe (2018), cheetah (2019), and African wild dog (2021). Majete is now Malawi's only Big Five reserve and the country's premier wildlife destination.

So successful has the recovery been that since 2016, Majete has been able to supply over 1,100 animals to help restore other parks across Malawi — including 150 elephants as part of the historic 500 Elephants translocation in 2017. The reserve's 2024 aerial census recorded more than 12,400 large herbivores alongside primates, predators, vultures, and ground hornbills.

Today, Majete is a living example of how sound management, community partnership, and sustained investment can transform a landscape. Zebra, buffalo, kudu, sable antelope, impala, and warthog roam the miombo woodlands and granite hills. Elephants bathe in the Shire River while hippos wallow nearby. Lions, leopards, and cheetahs have reclaimed their role as apex predators. And over 400 bird species — including Livingstone's flycatcher, Boehm's bee-eater, and four species of vulture — fill the air.

700 km² Reserve Area
3,000+ Animals Reintroduced
17 Species Restored
400+ Bird Species
Big Five Malawi's Only Reserve

What is MajeteWatch?

MajeteWatch is a real-time wildlife detection and survey system built specifically for Majete Wildlife Reserve — and it works in a way that has never been done before in conservation monitoring.

Rather than traditional camera traps that store footage locally, MajeteWatch is built around a network of six PixCams EZ Streamer live streaming systems deployed across the reserve. Each unit is entirely off-grid — powered by solar panels and batteries — and transmits a continuous live video feed to YouTube via Starlink satellite internet. This means the cameras are streaming 24 hours a day, 7 days a week, from deep inside one of Africa's most remote wildlife reserves.

What makes MajeteWatch truly unique is that the AI detection happens on the live stream itself. A Camect AI hub analyses the incoming video feeds in real time, detecting animal movements, people, and vehicles as they appear on screen. When a detection is triggered, MajeteWatch automatically saves a timestamped snapshot image, logs the camera location and detected species, and displays it on this dashboard within seconds. No footage needs to be physically collected — the entire pipeline from detection to public dashboard is fully automated and continuous.

Combining off-grid solar and satellite live streaming with real-time AI analysis of that stream for wildlife detection is a first of its kind in conservation technology — and it is happening right now at Majete Wildlife Reserve in Malawi.

MajeteWatch is a PixCams, Inc. initiative developed in partnership with the Majete Wildlife Reserve to advance conservation technology in Africa.

How It Works

MajeteWatch integrates several cutting-edge technologies into a single unified platform running on dedicated hardware at the reserve.

📡
PixCams EZ Streamer — Off-Grid Live Streaming
Each of the six camera stations is built around a PixCams EZ Streamer encoder — a compact unit that captures video and pushes a continuous live stream to YouTube via Starlink satellite internet. Every unit runs entirely off-grid on solar panels and batteries, making it deployable anywhere in the African bush regardless of infrastructure. The live streams run 24/7, giving both the AI system and the public a continuous real-time view of the reserve.
🎯
Camect AI — Live Stream Detection
A Camect AI hub receives all six live streams simultaneously and analyses them in real time using computer vision. When an animal, person, or vehicle appears on screen, Camect classifies the detection within seconds and triggers an alert. MajeteWatch listens to these alerts continuously and automatically logs every event — making AI-powered analysis of a live satellite stream the core of the system, a first in wildlife conservation monitoring.
🤖
MegaDetector AI
Every snapshot image is processed nightly by MegaDetector v5a — a deep learning model developed by Microsoft AI for Earth specifically for wildlife camera trap analysis. MegaDetector classifies each image as Animal, Person, or Vehicle with a confidence score, providing a second layer of AI verification on top of the Camect system. Over time, these labels are used to build a custom African wildlife classifier trained on real Majete images.
🐦
BirdWeather & BirdNET
Each camera station is equipped with an audio classifier powered by BirdNET — a neural network developed by Cornell Lab of Ornithology and Chemnitz University of Technology that identifies bird species from sound. Detections are streamed to BirdWeather, providing species-level bird identification across all five stations with confidence scores and audio recordings.
☀️
Solar & Satellite — Fully Off-Grid
Every camera station operates entirely independently of mains power and terrestrial internet infrastructure. Solar panels charge on-site batteries that power the cameras and streaming equipment around the clock. Starlink satellite dishes provide high-speed internet deep inside the reserve — making it possible to live stream and run AI analysis from locations that would otherwise be completely unreachable.
🗄
Detection Database
Every detection — mammal or bird — is stored in a searchable database with timestamp, camera location, species label, snapshot image, and AI confidence score. The database grows continuously and powers the calendar, charts, species breakdowns, and monthly survey reports visible on this dashboard.
🧬
Training Data Collection
MajeteWatch is actively building a labeled dataset of African wildlife images specific to Majete. Researchers can tag each detection with the correct species, creating training data for a future custom AI model that will identify individual species — elephant, lion, leopard, zebra, impala, and more — directly from camera trap images.

Why This Matters

Traditional wildlife surveys are expensive, time-consuming, and can only capture a snapshot of animal activity at the time a ranger or researcher is present. Automated camera trap monitoring changes this fundamentally — it provides continuous, objective, timestamped data across multiple locations simultaneously, without disturbing the animals.

For a reserve like Majete, where the recovery of predator and prey populations must be carefully monitored, this data is invaluable. Activity patterns reveal when and where animals are most active, which areas are being used by which species, whether nocturnal species are present, and whether human intrusion events require attention. Over time, trends in detection frequency provide early warning of population changes.

The bird audio detection adds another dimension — Majete's extraordinary diversity of over 400 bird species is an indicator of ecosystem health. Tracking which species are present, in what numbers, and at what times of year provides valuable data for biodiversity monitoring and research.

MajeteWatch makes all of this data publicly accessible in real time — connecting people around the world to the daily rhythms of one of Africa's most important conservation areas, and demonstrating what is possible when technology is put in service of wildlife protection.

Real-Time Conservation Alerts

Beyond passive monitoring, MajeteWatch includes a real-time alert system that notifies conservation staff and rangers the moment a critical event is detected — giving the team on the ground the ability to respond immediately, anywhere in the reserve, at any time of day or night.

🚨
Anti-Poaching Alerts
When a person or vehicle is detected inside the reserve, an immediate email alert is dispatched to rangers and conservation staff — including a timestamped snapshot image from the camera that triggered the detection, the camera location, and the exact time of the event. This gives rangers real-time intelligence to intercept potential poaching activity before it escalates. Alerts can be configured to activate only during high-risk hours, such as overnight, to focus attention when it matters most.
🦁
Rare Species Alerts
MajeteWatch can be configured to send instant alerts when specific rare or high-value species are detected — such as black rhino, cheetah, African wild dog, or pangolin. For species that are critically endangered or have been recently reintroduced to the reserve, knowing immediately when and where they have been sighted is invaluable for population monitoring, welfare checks, and rapid response if an animal appears distressed or in danger.
📸
Photo Evidence
Every alert email includes the full 1440p snapshot image captured at the moment of detection — providing immediate visual evidence of the event. Images are also permanently stored in the MajeteWatch database with their timestamp, camera location, and AI classification, creating an unbroken chain of photographic evidence that can support anti-poaching prosecutions, scientific research, and conservation reporting.
24/7 Autonomous Monitoring
The alert system operates entirely autonomously — no human needs to be watching the cameras for an alert to fire. Whether it is 2 AM on a Tuesday or a public holiday, if a person enters a restricted area or a rare species is sighted, the system detects it, logs it, photographs it, and notifies the team within seconds. This extends the effective reach of Majete's ranger force across the entire 700 km² reserve around the clock.

About PixCams, Inc.

PixCams, Inc. is a technology company specialising in AI-powered camera systems and wildlife monitoring solutions. MajeteWatch was designed, built, and is maintained by PixCams as part of our commitment to applying advanced technology to conservation challenges in Africa and beyond.