Project Scarecrow is JVR Software’s concept for an integrated, non-lethal-first robotic security layer built around the realities of South African farms.
Public concept release: 14 September 2026 · No weapons in v1 · No performance claims before field validation
Conceptual farm map showing fixed sensing, a local gateway, a supervised robotic patrol route and optional aerial observation.
Fences, cameras and alarms matter, but large properties still create blind spots and repetitive patrol work. Scarecrow’s thesis is simple: let machines perform consistent observation and verification, while trained humans make the decisions that carry legal and human consequences.
These are AfriForum figures, not official SAPS statistics. Definitions and datasets differ, so we cite the source and do not present the figures as complete national ground truth. Read the source.
That official figure replaces the unsupported ±68,000 figure in early Scarecrow drafts. It is a market context, not a claim that every farm needs or can afford this system.
Statistics South Africa, Census of Commercial Agriculture 2017.
The differentiator is the integration: fixed sensing, mobile verification, optional aerial observation, local decision support and a documented path to human response.
Site-appropriate thermal, optical, acoustic and perimeter inputs. The final sensor mix depends on a farm survey and legal review.
Rugged quadruped platforms are being evaluated for repetitive routes, remote inspection and live visual verification.
Local compute is intended to combine events, retain essential operation when connectivity fails and limit unnecessary data leaving the property.
Drones may extend visibility where lawful, licensed and operationally appropriate. Autonomous aircraft are outside the current v1 promise.
Clear alerts, live evidence and simple actions: review, communicate, call response and record the incident.
Scarecrow is designed to work with farmers, lawful security providers and emergency services—not replace their judgement or authority.
The prototype code accepts authenticated synthetic or compatible test events, stores structured evidence locally, opens review incidents and records human decisions. Ten automated tests exercise the current software boundary; this is not a continuously operating farm deployment.
The public demo uses synthetic events and sends no farm information anywhere. It shows how an uncertain signal becomes a human-reviewed incident.
No live robot or farm sensor is connected yet. That is the next controlled-pilot milestone.
Project Scarecrow is not a licence to use force, not a replacement for law enforcement and not a claim of regulatory approval.
The internal legal, safety, privacy, incident-response and risk documents are working drafts for review by qualified professionals. They are not published as approved policies.
The system is intended to reduce uncertainty and help lawful responders act with better information—not to make irreversible decisions on its own.
Selected fixed sensors and, later, supervised mobile patrols would watch agreed routes and critical areas.
The lab gateway already supports human review; real camera and mobile-platform verification still require selected devices.
Lights, sirens or voice warnings may be tested only after legal, safety and site approval. No output is connected today.
The intended result is better evidence for humans who communicate and activate the farm’s agreed response procedure.
Dates and performance targets will only be committed after funding, vendor selection and partner capacity are confirmed.
The v0.2 gateway, simulator and ten automated tests establish the software workflow. They do not prove real hardware or field performance.
Select a non-actuating camera, alarm or sensor and measure authenticated intake, health, latency and loss-of-link behaviour.
Run a non-lethal-only pilot on one to three farms with baselines, incident logging, operator training and independent review.
Publish measured results, refine service and pricing, complete professional legal/compliance work and decide whether scale is justified.
Our press room preserves dated releases and distinguishes documented facts, open questions and conclusions.
The documented quotation exchange, later public material and the clarification JVR is seeking.
We are looking for partners who can turn a careful concept into field evidence and a supportable South African service.
Representative sites willing to help define routes, blind spots, practical constraints and meaningful success measures.
Patient pre-seed or strategic capital for hardware, integration, testing, insurance and professional compliance work.
PSiRA-aware partners who can define lawful monitoring, escalation and response workflows.
OEMs, local distributors, ruggedisation specialists, camera/sensor suppliers and integration engineers.
South African counsel, POPIA specialists, safety engineers and insurers willing to challenge the assumptions.
Universities, robotics labs and security researchers for evaluation, datasets, red-teaming and independent evidence.
Confidential detail is available selectively. The public site describes the purpose and development route. Technical designs, financial assumptions and operating procedures are shared only after fit is established and appropriate confidentiality is in place.
Request a briefingNo. It is a documented concept at partner-discovery stage. Hardware platforms exist, but the integrated Scarecrow product has not yet completed prototype and farm-pilot validation.
No. The v1 concept is non-lethal first, with observation, verification and deterrence. Human beings remain responsible for escalation and response. No lethal capability is offered in v1.
No. The project has identified relevant regulatory areas, but final service design and each deployment require qualified advice, appropriate partners and current approvals where applicable.
There is no final price list. Cost depends on site size, routes, terrain, sensor coverage, hardware, connectivity and monitoring. Vendor quotes and pilot economics must come before a reliable commercial price.
A mobile robot is one component. Scarecrow is the proposed integration of sensing, local compute, alerts, operating rules, human response, maintenance, privacy and safety into a farm-specific system.
The project is vendor-neutral and farm-specific. It focuses on a coordinated layer, local operation, explicit non-lethal governance, South African partner integration and a measured pilot method rather than claiming that a single imported platform is a finished security service.
The project library turns the underlying research and working documents into focused, readable public briefs. Sensitive layouts, procedures, credentials and unverified commercial assumptions remain private.
The layered architecture, event lifecycle, device-adapter boundary, current software and engineering roadmap.
Open technology brief →Site discovery, baseline, survey, controlled tests, measurements, responsibilities and acceptance gates.
Open pilot programme →Human authority, safe failure, privacy, cybersecurity, incident governance and the South African operating context.
Open governance framework →Verified context, withheld claims, evidence labels, public sources and the dataset a pilot must produce.
Open evidence register →The capabilities, due-diligence material, collaboration structures and milestones needed to close the real gaps.
Open partner brief →Requirements, implementation status, boundaries and the current controlled-site acceptance standard.
Open public specification →Submitting a form is an expression of interest, not a contract, investment offer, product order or promise of acceptance into a pilot.
We link the main public sources used for market context and regulatory framing. Their inclusion does not imply endorsement or approval of Project Scarecrow.
