• 151

    Projects submitted

  • 109

    Published projects

  • 47

    Comments left

  • 66

    Number of votes

  • 52

    days until voting ends

← Back to list

IAsmin from ARCA: AI for Climate Resilience on the Amazon Coast

  • Project manager on customer side

    Jessica Wandscheer

    Rare Brazil

    Manager, Development & Engagement

  • Category

  • Nomination

  • Goals

    1. Enable smallholder aquaculture and fishing communities on the Brazilian Amazon Coast (Pará and Maranhão states) to access timely, actionable environmental intelligence - reducing their exposure to climate change such as tide peak hour changes, flooding, salinity crashes, and thermal stress events.

    2. Deliver insights through IAsmin - a WhatsApp-based conversational AI agent that translates complex satellite data into plain-language audio alerts in Portuguese: designed for communities with low digital literacy and limited connectivity.

    3. Establish a climate governance model that links municipal governments, extractive reserves (RESEX), academic institutions (LAPMAR/UFPA), NGOs (Rare Brazil), and the private sector (Accenture) around shared environmental monitoring infrastructure.

  • Timeline

    march, 2025 — november, 2026
  • Project completion year

    2026

  • Project scale

    40000 subscribers
  • Results

    Early monitoring results show that IASMIN is delivering practical value to oyster-farming families in the Mãe Grande de Curuçá Extractive Reserve, where the pilot was launched. In a September 2026 assessment conducted in Lauro Sodré, all respondents reported using IASMIN to plan their working day and said they intend to continue using and recommend the solution. 83% reported gains in time and planning and 67% highlighted the ability to plan further ahead.

    All respondents said the tidal forecasts matched conditions observed in the field, demonstrating IASMIN’s perceived accuracy and its complementarity with traditional knowledge. The responses also showed that information is already reaching people beyond direct users through family and community sharing.

    These preliminary findings demonstrate early operational value and a promising contribution to climate resilience. Longer-term effects on production and income will be monitored as the solution evolves.

  • Project uniqueness

    IASMIN is a community-centered climate intelligence solution that combines satellite Earth Observation, on-orbit AI processing, IoT sensor data and generative AI to turn complex environmental data into hyperlocal, actionable insights for traditional fishing communities. Its uniqueness lies in bridging advanced technology with radical accessibility: delivering short, natural-language audio forecasts in Portuguese via WhatsApp - a familiar channel reaching 98% of Brazilians, with no app, login or digital literacy barrier.Launched in 2025 with 15 families in Curuçá, the pilot demonstrated autonomous adoption, as families spontaneously shared alerts with neighbours. Its hyperlocal tide predictions also outperformed the official Belém tide table in local accuracy. Now scaling towards 10,000+ families in 2026, IASMIN turns satellite and sensor data into practical knowledge, helping extractivist communities move decisions, anticipate climate conditions and strengthen local resilience.

  • Software used

    • In-situ sensors: LAPMAR/UFPA in-water IoT sensors (LoRa, NB-IoT); inter-satellite link (ISL) for sensor-triggered satellite tasking.

    • Development: Claude Code, GitHub Copilot (AI-assisted pipeline development); GitHub, Azure cloud, DevOps for sprint-based delivery.

    • Communication: WhatsApp Business API via BotMaker platform for message delivery; Conversational Orchestrator for intent routing and alert triggering.

    • AI/LLM: Large language model for data-to-insight translation (Portuguese-language alert generation), Text-to-Speech (TTS) for audio message production.

    • Processing: Google Earth Engine (cloud-based EO processing), Python-based ETL pipelines, ML models for indicator derivation (turbidity, chlorophyll-a, flood extent, mangrove NDVI).

    • Satellite data: Sentinel-1 (SAR, cloud-penetrating), Sentinel-2 (multispectral, 10–20m), Sentinel-3 (thermal/multispectral, 300m), Landsat, CHIRPS rainfall accumulation, ACCENTURE-1 hyperspectral (4.75m, 442 bands).

  • Solution from the Global CIO catalog

    The project does not use solutions from the Global CIO catalog

  • Project implementation complexity

    • Scientific and analytical complexity: Transforming raw environmental observations into meaningful indicators and forecasts. This requires integrating oceanographic and environmental expertise, data modelling, validation and the interpretation of uncertainty — particularly in dynamic estuarine ecosystems.

    • AI and technology architecture: Building cloud-based data pipelines that can process and connect these sources, generate relevant insights and translate them into concise, understandable messages. The architecture must also be reliable and scalable as the solution expands to new territories and communities.

    • Multi-stakeholder implementation: Coordinating scientific institutions, technology partners, government actors and community organizations, while aligning technical development with local knowledge, field validation, operational requirements and long-term ownership.

  • Description

    IASMIN: Making Climate Intelligence Accessible to Amazon Communities

    The Amazon Coast of Brazil is home to over 80,000 people across 500+ communities and 12+ federal marine extractive reserves. Extractivist communities and smallholder aquaculture farmers depend directly on estuarine ecosystems for their livelihoods. Yet climate change is making these environments increasingly unpredictable, with shifting tidal patterns disrupting production cycles and threatening local incomes.

    IASMIN, developed under ARCA (Alliance for Amazon Climate Resilience), transforms environmental data into actionable climate intelligence for these communities. Created through a collaboration between Accenture, Rare Brazil and LAPMAR/UFPA, the solution combines university-led IoT development, cloud infrastructure and AI capabilities delivered by Accenture, and community-centered design.

    LAPMAR/UFPA contributes scientific expertise and in-water monitoring infrastructure, including environmental sensors that capture local conditions. Accenture acts as the technology delivery partner, developing the cloud architecture, data modelling and AI capabilities that transform sensor readings and public datasets into meaningful forecasts and insights. These are delivered through short, natural-language audio messages in Portuguese via WhatsApp — a familiar, accessible channel that requires no app installation, login or specialized digital literacy. This voice-first approach makes advanced climate intelligence available to communities often underserved by conventional digital solutions and lacking access to territorialized data.

    Officially launched at COP30 in November 2025, IASMIN's initial pilot reached 15 families in Curuçá, demonstrating organic adoption as families began sharing alerts with neighbours. Its roadmap targets 10,000 families in 2026 and over 44,000 across the Amazon Coast by 2028.

    In 2026, IASMIN's intelligence capabilities are being expanded through a 16-week Earth Observation pilot. Accenture leads the development of the satellite data modelling and AI pipeline, integrating Sentinel, Landsat, CHIRPS and ACCENTURE-1 data with existing sensor readings. This additional layer enables new environmental indicators, including water turbidity, chlorophyll-a, surface temperature, flood extent and mangrove health, enriching the hyperlocal intelligence available to communities.

    A defining innovation is the translation of complex environmental data into locally meaningful guidance. Community field research informs a four-layer framework — signal, indicator, decision relevance and message — ensuring that technical information reflects the language, knowledge and decision-making needs of local communities.

    IASMIN is part of a broader climate resilience strategy led by Rare Brazil, connecting local adaptation needs with environmental evidence and pathways to international climate finance. Within this strategy, IASMIN contributes territorialized, trustworthy and actionable insights.

  • Project geography

    IASMIN was initially piloted in Curuçá, on the Amazon Coast of Pará state, Brazil. In 2026, the project is expanding its geographical coverage through the integration of Earth Observation (EO) data, extending its reach to other coastal municipalities, including Soure, Bragança and Augusto Corrêa, across the Salgado Paraense and Bragantina microregions — approximately 400 km of coastline.

    By 2028, the ambition is to extend IASMIN across the full Amazon Coast, spanning the states of Pará and Maranhão.

  • Additional presentations

Comment

Only authorized users can comment.
We suggest you or register.

  • Customer

    Rare Brazil

  • IT provider

    Accenture

We use cookies for analytical purposes and to deliver you the best experience with our website. Continuing to the site, you agree to the Cookie Policy.