Merge pull request #118 from Catrobat/AIOTSonline-patch-2

Update 2026.md

Author: AIOTSonline

pages/development/google-summer-of-code/2026.md

@@ -311,6 +311,154 @@ The architecture will allow reusable rocket parts, physics presets, and AR exper
 
 <hr>
 
+### Blockchain-Based Ethical Governance for IoT Care Systems
+
+350 Hours  
+{: .label .label-blue }
+
+{: .highlight }
+> **Required Skills**: Strong programming fundamentals, smart contract development, blockchain architecture understanding, API design and system integration, security and access-control concepts
+> **Possible Mentors**: Garima Jain, Supreeth Kumar M
+> **Desirable Skills**: Privacy-by-design principles, cryptographic hashing, regulatory-aware system design  
+> **Expected Outcome**: A complete blockchain-based consent and governance framework for IoT care systems  
+> **Difficulty level**: Advanced  
+> **Project Size**: Large
+
+Description
+
+This project focuses on designing and implementing a blockchain-based governance and consent framework for ethical IoT-based care monitoring systems. The objective is to ensure trust, transparency, data ownership, and immutable consent management for sensitive care-related data generated by IoT devices used in assisted living environments.
+
+Rather than storing health data on-chain, the blockchain will act as a trust and audit layer, recording consent decisions, access approvals, role assignments, and accountability events.
+
+Problem Context
+
+IoT-based care systems face critical challenges:
+
+- Consent is often implicit, unclear, or changeable without traceability  
+- Care data access decisions are difficult to audit  
+- Families, caregivers, and supervisors rely on centralized systems with limited transparency  
+- Ethical compliance relies heavily on documentation rather than system-level enforcement  
+
+This project addresses these issues by embedding ethical governance directly into system architecture using blockchain.
+
+Technical Scope
+
+**A. Governance Model Design**
+- Definition of care-related roles (patient, caregiver, supervisor, relative)  
+- Consent lifecycle modeling (grant, update, revoke)  
+- Human-in-the-loop approval workflows  
+
+**B. Blockchain Layer**
+- Smart contracts for:
+  - Consent registration and revocation  
+  - Role-based access authorization  
+  - Event logging for care-related decisions  
+- Immutable audit trail for:
+  - Who approved what  
+  - When access was granted or revoked  
+  - Which role initiated the action  
+
+**C. Off-Chain / On-Chain Architecture**
+- Sensitive IoT data stored off-chain  
+- Cryptographic hashes and metadata stored on-chain  
+- Blockchain used purely for verification, not data storage  
+
+**D. Integration Interface**
+- APIs to connect IoT systems with the blockchain layer  
+- Verification endpoints for access checks  
+- Read-only audit views for compliance and evaluation  
+
+Expected Outcome
+
+At the end of the project, the contributor will deliver:
+
+- A complete blockchain-based consent and governance framework  
+- Deployed smart contracts implementing ethical access control  
+- API layer enabling IoT system integration  
+- Demonstrable immutable audit trail  
+- System architecture documentation and threat analysis  
+
+The result is a production-relevant governance layer, not a theoretical blockchain demo.
+
+<hr>
+
+### Gemini-API–Powered Intelligent Care Assistant
+
+350 Hours  
+{: .label .label-blue }
+
+{: .highlight }
+> **Required Skills**: Backend development, API integration, prompt engineering for structured systems, data processing and normalization, system-level reasoning  
+> **Desirable Skills**: Human-centered AI design, ethical AI concepts, evaluation of AI outputs  
+> **Possible Mentors**: Supreeth Kumar M, Atharva Prashant Joshi
+> **Expected Outcome**: An AI-powered, ethically governed intelligent care assistant built on Gemini API  
+> **Difficulty level**: Intermediate to Advanced  
+> **Project Size**: Large
+
+Description
+
+This project aims to build an AI-powered intelligent care assistant using the Gemini API to support caregivers, supervisors, and families by converting raw IoT activity signals into context-aware insights, summaries, and alerts—while maintaining ethical, permission-based access. The focus is on responsible AI usage, ensuring that AI augments human decision-making rather than replacing it.
+
+The assistant acts as an interpretation and explanation layer over existing IoT-based care systems, not as a surveillance or diagnostic system.
+
+Problem Context
+
+IoT care systems generate large volumes of low-level signals such as:
+
+- Activity logs  
+- Time-based events  
+- Movement patterns  
+- Routine confirmations  
+
+These signals are difficult to interpret meaningfully and ethically in real time. Manual monitoring often leads to caregiver overload, missed anomalies, and increased anxiety among family members. This project introduces an AI interpretation layer that summarizes and contextualizes data while preserving human oversight.
+
+Technical Scope
+
+**A. Data Interpretation Layer**
+- Structured ingestion of non-invasive IoT activity data  
+- Time-windowed summaries of daily routines  
+- Detection of deviations from normal patterns  
+
+**B. Gemini API Integration**
+- Natural-language summaries of patient routines  
+- Context-aware explanations for alerts (why an alert was triggered)  
+- Ethical prompt design to avoid medical diagnosis or inference  
+- Role-aware output filtering (different outputs for caregivers, supervisors, and relatives)  
+
+**C. Human-in-the-Loop Controls**
+- AI outputs require supervisor validation before escalation  
+- Confidence indicators and uncertainty explanations  
+- Manual override and feedback loop for continuous improvement  
+
+**D. Responsible AI Safeguards**
+- Prompt constraints and system instructions  
+- No medical diagnosis generation  
+- Explainability-first responses  
+- Logging and review of AI outputs  
+
+**E. Application & System Integration**
+- Integration with a secure care monitoring application featuring role-restricted dashboards  
+- Support for voice-assisted interactions to improve accessibility for elderly users  
+- Backend services handling ingestion, summarization, and alert generation  
+- Secure authentication and role-based access to AI-generated insights  
+
+Expected Outcome
+
+By the end of the project, the contributor will deliver:
+
+- An AI-powered intelligent care assistant service  
+- Gemini API–based summarization and explanation engine  
+- Ethical prompt and output governance framework  
+- Role-based AI response filtering  
+- Demonstration of AI-assisted, human-approved alerts  
+- Complete documentation and evaluation report  
+
+The outcome demonstrates applied, responsible AI in a real-world care context, not chatbot experimentation.
+
+<hr>
+
+
+
 ### AR Gravity & Planetary Physics Simulator
 
 175 Hours