← Investigations

Investigation 01 / Active — 2021 → Present

Kaavalar.

Protecting sandalwood ecosystems through indigenous security infrastructure.

Type
Investigation · Ecology · Hardware
Timeline
2021 — Present
Role
Lead Researcher · Field Engineer
Location
Tumkur, Karnataka
Status
Active investigation
Close-up of sandalwood tree bark in soft natural light.
KAV.01 / Sandalwood canopy, Tumkur district

Themes

EcologyStewardshipIndigenous KnowledgeIoTCommunity InfrastructureConservation

Kaavalar began as a question:

How might technology protect forests without replacing the communities that have protected them for generations?

Developed through field research with sandalwood farmers in Karnataka, the project investigates indigenous security systems, ecological stewardship, sensor networks, and distributed monitoring infrastructures designed to safeguard one of India's most valuable and vulnerable ecological resources.

This is not simply a technology project. It is an inquiry into guardianship.

The Problem

A tree worth more than the systems meant to protect it.

Sandalwood cultivation represents a vital economic resource for thousands of farmers across Southern India. Yet the value of sandalwood has created a parallel ecosystem of theft, illegal extraction, and violence.

Farmers are often forced to rely on expensive barriers, guards, or reactive security measures that fail to address the root problem.

Kaavalar investigates how long-term, affordable, community-owned systems might provide resilience without increasing dependence on external actors.

Stakeholder Ecosystem

The tree is held by more hands than it looks.

Sandalwood does not stand alone. It sits inside an ecosystem of farmers, cooperatives, forest officers, traders, processing units, regional markets, customary guardians, and the families whose long-term livelihoods are tied to the canopy.

Early in the project we mapped this network deliberately — not as an org chart, but as a field of overlapping responsibilities. Every later design decision was tested against the question: whose hands does this strengthen, and whose does it sideline?

  • ·Sandalwood farmers and family stewards
  • ·Local cooperatives and farmer producer organisations
  • ·Karnataka State Forest Department
  • ·Customary guardians and night-watch networks
  • ·Traders, auctioneers, and processing units
  • ·Researchers, NGOs, and policy partners

Farmer Stories

What the trees are worth, told by the people who guard them.

"I planted these trees the year my eldest daughter was born. They will outlive me. Anything that protects them protects her too."

— Farmer, Tumkur district · 2023

"The guards we hire are expensive. The fence is expensive. The fear is the most expensive thing of all."

— Cooperative member · 2023

"If the device only tells the police, it is not ours. If it tells us first, it becomes part of the village."

— Senior guardian · 2024

Security Fatigue

The hidden cost of guarding a tree for thirty years.

Theft is the visible problem. The invisible problem is fatigue. A sandalwood tree matures over decades — and across those decades a family carries the constant, low-frequency cost of vigilance: paying for guards, repairing fences, sleeping lightly, losing trust in neighbours, deciding whether to plant another generation of trees at all.

Any technology that adds to this load fails, no matter how well it performs on paper.

Kaavalar is therefore evaluated against a quieter metric: does it reduce the daily emotional and financial weight of guardianship for the people who already carry it?

Field Research

Listening before building.

The project began through conversations with sandalwood farmers, field visits, stakeholder mapping, market analysis, and technical feasibility studies.

Research focused on

  • ·Tumkur district, Karnataka
  • ·Sandalwood cultivation practices
  • ·Existing protection systems
  • ·Forest department policies
  • ·Farmer experiences with theft
  • ·Long-term sustainability requirements

The objective was not simply to build technology. The objective was to understand the ecosystem surrounding the tree.

System Design

A four-layer protection architecture.

L.01 MOTION GUARDL.02 COMMUNITYL.03 TREE RINGCHAIN

Fig. KAV.A / Layered architecture — schematic

  1. Layer 01

    Motion Guard

    Perimeter-based motion detection integrated into existing solar fencing systems.

  2. Layer 02

    Community Surveillance

    Partnerships with local security ecosystems and traditional protection practices.

  3. Layer 03

    Tree Ring

    Low-power motion sensing infrastructure installed directly on trees.

  4. Layer 04

    Chain Node

    Embedded GPS and IoT infrastructure distributed throughout the plantation.

Together these layers form a resilient, distributed security architecture rather than a single point of failure.

Chain Node

The final line of defence.

At the core of Kaavalar sits Chain Node.

A distributed network of GPS-enabled sensor modules embedded throughout sandalwood plantations.

Rather than relying on one central device, the system operates as a mesh of interconnected nodes capable of tracking movement, detecting tampering, and maintaining awareness across large areas.

The architecture prioritizes

  • ·Low power consumption
  • ·Long operational lifespan
  • ·Affordability
  • ·Scalability
  • ·Community ownership

Prototyping

Building in the field.

The project moved through multiple rounds of experimentation involving:

  • ·Raspberry Pi systems
  • ·GSM communication modules
  • ·NodeMCU ESP8266
  • ·GPS NEO 6M modules
  • ·Arduino-based testing
  • ·Rapid prototyping

Every iteration was informed by constraints encountered on the ground.

Power availability.Cost.Maintenance.Environmental conditions.

These constraints shaped the design.

Systems Thinking — Iceberg Model

Theft is the event. The system is everything beneath it.

  1. 01 · Events

    Theft incidents

    Visible occurrences — a tree felled, a fence cut. Most security spending stops here.

  2. 02 · Patterns

    Price spikes · seasonal migration · gaps in presence

    Repeating conditions that make events more or less likely across months and years.

  3. 03 · Structures

    Licensing · markets · access to credit · policing

    Institutional shapes that produce the patterns. Slow to change, but where leverage lives.

  4. 04 · Mental Models

    The tree as commodity · as kin · as inheritance

    The deepest layer — how stakeholders imagine the tree itself. Interventions here are the most durable.

Kaavalar is positioned across all four layers — sensors at the event layer, cooperatives at the pattern layer, policy work at the structural layer, and farmer narratives at the level of mental models.

Design Process

Iterating with the field, not for it.

Each revision of Kaavalar was reviewed with farmers, cooperative members, and forest officers before becoming the basis for the next prototype. The process is closer to ethnographic research than product design — slow, dialogic, and willing to throw away a working prototype if it solves the wrong problem.

  • ·Stakeholder mapping and conversation
  • ·On-site observation and transects
  • ·Constraint capture — power, cost, terrain, maintenance
  • ·Bench prototyping in Bangalore
  • ·Field deployment in Tumkur
  • ·Review with farmers and cooperatives
  • ·Refinement and the next revision

The Kaavalar Platform

A platform held by the community, not above it.

Farmer Layer

A lightweight Kannada-first interface — daily status, alerts, and acknowledgements over web and SMS.

Cooperative Layer

Shared visibility across member plantations. Coordinates response, maintenance rotas, and learning between farms.

Steward Layer

Aggregated, anonymised insight for forest officers, researchers, and policy partners — released with farmer consent only.

Keys, alerts, and decisions flow outward from the farmer — not inward from a central authority. The platform is designed so the people closest to the trees see things first.

Future Directions

Where Kaavalar is heading.

The next phase of the project moves from prototype to participatory infrastructure — longer field deployments, an open four-layer reference architecture, a farmer-facing dashboard, and a small policy brief written with cooperatives rather than about them.

Beyond sandalwood, the same model is being explored for other high-value crops and for ecological monitoring more broadly — wherever value, vulnerability, and long stewardship intersect.

Chapter — Field documentation

A pair of cylindrical ambisonic microphones with perforated mesh grilles on a warm orange reflective surface.
FLD.001
Tree-ring sensor, prototype rev. 3 — bark-mounted housing.
Hand-drawn ledger of food supply chain nodes on aged paper.
FLD.002
Stakeholder mapping session — Tumkur cooperative, 2023.
Warm terracotta interior wall with timber screen casting linear shadows.
FLD.003
Solar fence integration — Motion Guard layer.
Misty sandalwood forest at dawn in the Western Ghats.
FLD.004
Three-hour transect with senior guardian.
A hand holding a handheld field recorder at the edge of a Bangalore lake, overcast sky and acacia growth.
FLD.005
Locally machined teak housing for Chain Node.
A Bangalore lake at dawn — marsh grasses, still water, distant apartment towers under overcast sky.
FLD.006
Topographic overlay — four-year change detection.

Impact

Stewardship infrastructure.

Kaavalar imagines technology as a supporting layer beneath existing ecological knowledge.

The goal is not surveillance. The goal is stewardship.

By reducing theft, increasing traceability, and strengthening farmer agency, the project contributes toward:

  • ·Economic resilience
  • ·Ecological protection
  • ·Biodiversity preservation
  • ·Community empowerment
  • ·Long-term sustainable cultivation

Field note / 14 March 2024

"The forest has its own clock. The sensors record not just data, but the rhythm of the canopy. To protect the tree is to protect the community that guards it."

Archive

Field notes, maps, and prototypes.

A living archive of documents, conversations, build logs, and references — published as each layer of the project is released into the open.

Archive · Images

Tumkur — field documentation

Tumkur — field documentation

Archive · Images

Hardware iterations — workshop record

Hardware iterations — workshop record

Archive · Documents

Papers, reports, downloads

PaperComing Soon

Kaavalar — Research Brief (Vol. I)

Field methodology, stakeholder map, and early findings from Tumkur fieldwork.

Pending
PaperIn Progress

Four-Layer Architecture — Technical Whitepaper

System diagram, layer responsibilities, failure modes, and resilience analysis.

Pending
DocIn Progress

Chain Node — Build Notes (rev. 1 → rev. 4)

Bill of materials, mesh protocol notes, power budget, and enclosure drawings.

Pending
DocComing Soon

Farmer Conversations — Field Journal

Edited transcripts and notes from on-site interviews, 2022 — 2024.

Pending
PaperComing Soon

Mesh resilience under intermittent GSM

Measured behaviour of the node network across patchy rural connectivity.

Pending
PresentationComing Soon

Workshop deck — Participatory security systems

Teaching deck used with cooperatives, students, and policy partners.

Pending

Archive · Field notes

From the field

  1. 2024 · 03

    Transect with a senior guardian

    Three hours along the southern boundary. He reads the plantation by sound and absence — birds that should be there, footfalls that should not. The sensor network is, at best, a memory aid for what he already knows.

  2. 2023 · 11

    Cooperative meeting — Tumkur

    Eleven farmers, two forest officers, one cooperative manager. Consensus on the four-layer model only emerged once 'community surveillance' was named explicitly as a layer of equal weight to the hardware.

  3. 2023 · 06

    Security fatigue

    A recurring theme: farmers are exhausted by the cost and emotional load of guarding their own trees. Any new system has to lower that load, not add to it. This reframed our entire UX brief.

  4. 2022 · 09

    First plantation visit

    Walked the perimeter at dusk. Solar fence intact, but the back ten acres are effectively unguarded after dark. The geography of theft is not random — it tracks the gaps in human presence.

Archive · Research notes

Reading, thinking, marginalia

  1. 2024 · 05

    Iceberg model — applied to sandalwood theft

    Events (theft incidents) sit on top of patterns (price spikes, migration), patterns sit on structures (licensing, market access), structures sit on mental models (the tree as commodity vs. as kin). Interventions at the top of the iceberg are the least durable.

  2. 2024 · 02

    Reading — Elinor Ostrom on commons governance

    Re-reading Ostrom alongside Tumkur field notes. The eight design principles for stable commons map cleanly onto what farmers are already doing; our role is infrastructural, not organisational.

  3. 2023 · 08

    On 'participatory security'

    Notes towards a paper: distinguishing surveillance (extractive, top-down, opaque) from stewardship (reciprocal, distributed, legible). The same hardware can perform either role depending on who holds the keys.

Archive · Roadmap

What comes next

  1. Chain Node rev. 5 — field deployment

    In Progress

    Twelve nodes across two cooperating plantations, instrumented for a six-month resilience study.

  2. Public release — four-layer reference architecture

    Coming Soon

    Open documentation of the layered model so other cooperatives, forest officers, and researchers can adapt it.

  3. Kaavalar platform — farmer-facing dashboard

    In Progress

    Lightweight web and SMS interface designed with cooperatives for daily use, in Kannada and English.

  4. Policy brief — community-owned monitoring

    Coming Soon

    Short document for state forest departments outlining a participatory alternative to private guarding contracts.

  5. Extension — applying the model beyond sandalwood

    Coming Soon

    Exploratory work on adapting the four-layer architecture to other high-value, theft-prone crops and to ecological monitoring more broadly.