Scalable | Secure | Sustainable | Smarter Power for the Future
An Intelligent, Hybrid IoT Architecture for Next-Generation Energy Management
Kerala’s energy infrastructure is evolving — and with it comes the challenge of ensuring efficient, balanced, and intelligent power distribution across every region.
Our solution, the Smart Phase Balancing and Hybrid IoT Power Distribution System, redefines how transformers and homes communicate, coordinate, and optimize electricity delivery.
This project builds a self-learning, hybrid IoT architecture that connects every household node and transformer to a central intelligence layer — providing real-time visibility, load balancing, and automated control across both urban and rural grids.
It’s more than just monitoring — it’s about creating a power network that can sense, decide, and act.
⚡ Built for Scalability. Designed for Security. Engineered for Real-Time Performance.
Kerala’s power grid spans a complex mix of dense city networks, campuses, and rural terrains.
While urban regions are data-rich, rural and hilly areas still lack connectivity — resulting in phase imbalance, transformer overloads, and manual dependency.
Our system ensures that no matter the terrain, every node, every transformer, and every watt of energy is monitored, managed, and optimized.
This architecture directly supports KSEB’s vision for a Smart, Connected Kerala, integrating renewable energy and improving grid resilience.
| Feature | Traditional Grid | Our Smart Grid |
|---|---|---|
| Phase Monitoring | Manual, delayed | Real-time IoT sensing |
| Control | Manual switchovers | Automated balancing logic |
| Connectivity | Wired or single-channel | Hybrid (Cellular + LoRa + Wired) |
| Data Handling | Limited local logging | State-wide data ingestion & analytics |
| Scalability | Transformer-level | State-wide, modular, and scalable |
| Security | Minimal | TLS + AES + Role-Based Control |
🧠 This isn’t just digitalization — it’s intelligence built into every transformer.
Our primary goal is to design a modular, scalable, and cost-efficient hybrid IoT ecosystem that enables smart, automated, and secure energy distribution.
- Detect phase imbalance and transformer overloads automatically in real time.
- Ensure reliable communication between every home node, transformer, and control center using hybrid connectivity.
- Optimize energy flow and distribution efficiency with autonomous local decision-making at the edge.
- Deliver real-time dashboards and analytics for monitoring, alerts, and performance insights.
- Maintain data security and compliance through encryption, authentication, and audit tracking.
- Support urban, campus, and rural topologies under one unified architecture.
- Align with the Smart India Mission and KSEB’s modernization roadmap for sustainable grid transformation.
“Every transformer thinks. Every node communicates. Every decision optimizes the grid.”
Our design philosophy is grounded in three pillars:
- ⚙️ Modularity – Independent components (hardware, software, communication) that integrate seamlessly.
- 🧠 Intelligence – Edge analytics and centralized data insights for predictive and preventive action.
- 🔒 Security & Reliability – Multi-layer protection ensuring trusted and resilient energy communication.
Smart distribution starts at the transformer — but intelligence begins at every node.
Electricity distribution begins at the transformer and the household node — our design ensures that intelligence is embedded at the point of action.
The system operates on a distributed model where each hardware layer — home node, transformer PI, and gateway — contributes to real-time balance, control, and reporting.
Each connected home is equipped with a Smart Node Module, built around low-cost yet reliable microcontrollers that can measure, communicate, and act.
Key Components
- ⚡ Microcontroller: Arduino or ESP32 for real-time control and data handling.
- 🧲 Sensors: PZEM-004T or equivalent modules to measure current, voltage, and power factor.
- 🔄 3-Phase Relay Module: Dynamically switches household load between phases under PI command.
- 📡 Communication Interface: RS-485 (wired) or LoRa/Wi-Fi (wireless) for reliable connectivity.
Functions
- Constantly monitors power flow and sends updates to the transformer-level PI.
- Switches phases autonomously when instructed or during local instability.
- Can detect anomalies such as overcurrent or reverse feed from solar sources.
- Capable of limited edge processing — reducing unnecessary communication load.
🧠 Each home node acts as a “thinking endpoint” — small in size, big in intelligence.
At the center of each transformer sits a Raspberry Pi-based controller, the real-time decision-maker of the network.
Core Responsibilities
- Aggregates readings from 50–100 connected home nodes.
- Continuously evaluates phase balance, load deviation, and power quality.
- Executes control signals to nodes for instant phase reallocation.
- Communicates summarized reports to the Central Control Center (server).
Integrated Features
- Computation: Load-balancing algorithm for rapid phase switching.
- Preprocessing: Data aggregation and anomaly detection before transmission.
- Connectivity: Cellular IoT (NB-IoT / LTE-M) or LoRa mesh uplink.
- Metering: Uses modules like EM-10 for precise 3-phase monitoring.
- Local Storage: Temporary SQLite buffer for outage resilience.
“Where traditional transformers stop at distribution — ours begins with intelligence.”
In low-connectivity areas, a designated Gateway PI collects data from nearby PIs via LoRa and pushes it to the control center using a stronger cellular link.
- Acts as a local data hub for multiple transformers.
- Reduces SIM usage and overall communication cost.
- Ensures complete coverage, even in rural or hilly terrains.
Hardware built for field reality — rugged, scalable, and ready for automation.
The network is the spinal cord of this entire system — securely carrying commands, events, and data across mixed geographies.
Our design follows a Hybrid IoT Connectivity Model, ensuring every transformer remains online, regardless of infrastructure constraints.
- Communication: RS-485 wired between meters and PI.
- Why: Short distances, minimal noise, and near 100% uptime.
- Advantage: No interference, low cost, and highly accurate load reporting.
✅ Goal: Maximum reliability and accuracy in compact areas.
- Communication: RS-485 within each feeder + LoRa uplink to the central PI.
- Why: Reduces cabling mess across wide areas (100–500m).
- Advantage: Local wired stability combined with long-range wireless communication.
✅ Goal: Maintain feeder-level precision while covering larger distances.
- Communication: LoRa wireless nodes → Gateway PI → Control Center.
- Why: Long-range communication (up to 10 km), low power, solar-compatible.
- Advantage: Enables rural electrification without costly infrastructure.
✅ Goal: Reliable long-range telemetry for remote grids.
| Environment | Communication Mode | Range | Reliability | Cost | Ideal Use Case |
|---|---|---|---|---|---|
| Urban | RS-485 (Wired) | <100 m | ⭐⭐⭐⭐ | 💰 Low | Streets, dense housing |
| Campus | LoRa + RS-485 | 0.5–2 km | ⭐⭐⭐⭐ | 💰💰 Moderate | Institutions, feeders |
| Rural | LoRa → Gateway PI | 1–10 km | ⭐⭐⭐ | 💰 Very Low | Villages, farmlands |
🛰️ From cables to the cloud — our hybrid network ensures every watt is heard.
Beneath the field layer lies the digital intelligence stack — the system’s analytical brain that transforms raw energy readings into actionable insights.
Our software architecture is modular, secure, and horizontally scalable — designed to handle thousands of simultaneous PI connections while maintaining real-time responsiveness.
- Protocol: MQTT over TLS — lightweight and secure IoT communication.
- Message Broker: Apache Kafka — high-throughput stream ingestion for continuous PI data.
- Edge Preprocessing: PI filters redundant data to reduce transmission volume.
Kafka acts as the digital artery — buffering and streaming millions of readings without data loss.
- TimescaleDB: Optimized for time-series IoT data — fast reads and writes.
- PostgreSQL: Stores metadata, configurations, and user information.
- On-Premise Data Lake (HDFS / NFS): Retains historical records for audits and AI-based forecasting.
Every second of power history — preserved, queryable, and meaningful.
- Apache Storm: Real-time stream processor detecting anomalies, overloads, and phase imbalances.
- Apache Druid: Batch analytics engine performing trend and capacity studies.
- Visualization Layer: Grafana and Apache Superset for dashboards, alerts, and control views.
From milliseconds to months — every insight drives smarter decisions.
- Backend Framework: Python (FastAPI) + NGINX for secure routing.
- Frontend: React.js for responsive, modular dashboards.
- Visualization: Grafana for live status; Superset for analytics.
- Communication: REST and MQTT APIs for real-time control and data sync.
Control, visualize, and act — all from a single unified interface.
💬 From field data to decision dashboards — intelligence flows seamlessly through every layer.
In critical infrastructure like the power grid, security and reliability are non-negotiable.
Our architecture embeds protection, resilience, and compliance at every layer — from field hardware to analytics servers.
This ensures that all data remains authentic, encrypted, and tamper-proof, while operations continue seamlessly even during outages or network fluctuations.
| Layer | Security Mechanism | Purpose |
|---|---|---|
| Communication | TLS 1.3 Encryption | Secures data transmission between PI units and Control Center, preventing interception. |
| Storage | AES-256 Encryption | Protects data at rest within databases and logs, ensuring confidentiality and compliance. |
| Access Control | JWT + RBAC | Provides authenticated and role-based access for users and devices. |
| Network | VPN + Firewalls | Isolates the internal communication network, blocking unauthorized intrusion. |
| Governance | Audit Logging System | Tracks all configuration changes, access events, and control actions for traceability. |
Security is not an afterthought — it’s embedded by architecture.
- TLS 1.3: Prevents data theft or man-in-the-middle attacks during communication.
- AES-256: Ensures that even if storage is breached, data remains unreadable.
- JWT Authentication: Blocks unauthorized logins and protects sensitive control APIs.
- RBAC: Limits user operations — only authorized roles can trigger control actions.
- VPN & Firewalls: Shield the network backbone from external interference.
- Audit Logs: Enable forensic tracing and compliance with KSEB cybersecurity norms.
Together, these measures prevent breaches, spoofing, and unauthorized command injection, ensuring a trusted and resilient network for smart power control.
Security is complemented by robust reliability features that guarantee uptime and operational continuity:
| Mechanism | Description |
|---|---|
| Local Edge Autonomy | Each PI can operate independently when connectivity is lost — continuing to balance loads locally. |
| Data Buffering | Temporary local storage using SQLite ensures no data loss during transmission breaks. |
| Failover Servers | Backup servers in the Control Center replicate key services for uninterrupted operation. |
| Automated Backups | Scheduled daily and weekly backups protect against hardware failure. |
| Redundant Communication Channels | Multiple paths (Cellular, LoRa, Wired) ensure data flow continuity in any condition. |
🔁 The grid never sleeps — neither does its monitoring.
Our design aligns with key government and industry-grade security frameworks:
- KSEB IT & Data Protection Policies
- CERT-In Cybersecurity Guidelines for Critical Infrastructure
- ISO/IEC 27001 – Information Security Management
- National Smart Grid Mission (NSGM) Standards
By adhering to these, the system guarantees national-level cybersecurity readiness while maintaining local deployment feasibility.
Power you can trust — secured by design, resilient by architecture.
A truly smart system doesn’t just work — it monitors itself, repairs itself, and reports intelligently.
Our monitoring and reliability framework ensures that every device, every connection, and every process in the network stays healthy, responsive, and optimized 24×7.
We use open-source observability tools that continuously monitor hardware, communication links, and data flow in real time.
| Component | Tool Used | Purpose |
|---|---|---|
| System Metrics | Prometheus | Tracks CPU usage, memory, network latency, and ingestion rates. |
| Visualization | Grafana | Displays real-time dashboards for health, performance, and alerts. |
| Central Logging | ELK / EFK Stack | Aggregates logs from every Raspberry Pi, server, and API for quick fault diagnosis. |
| Network Health | Ping & MQTT Response Tracker | Measures packet loss, latency, and node availability. |
| Device Status | Heartbeat Protocol | Each PI sends periodic pings to confirm operational state. |
Continuous monitoring transforms potential failures into proactive fixes.
Real-time alerts are automatically triggered when anomalies occur — ensuring faster human response and system resilience.
| Event Type | Trigger | Response Mechanism |
|---|---|---|
| Transformer Overload | Voltage/Current Threshold Crossed | Alert to Dashboard + PI auto-balancing command |
| Communication Failure | No data heartbeat > 2 min | Local buffering + Email/SMS alert |
| Device Fault | Node unresponsive | Field engineer notification + log capture |
| Data Spike / Noise | Unusual input pattern | Temporary isolation + flagged for inspection |
Alerts can be sent to:
- Government / KSEB Engineers: via SMS or Email notifications.
- Dashboard Operators: via Grafana or Web Portal alerts.
- Field Maintenance Teams: via Telegram/Slack integrations.
🧠 We don’t wait for faults — we anticipate and prevent them.
Reliability is achieved through a combination of redundancy, fail-safes, and modularity.
| Feature | Description |
|---|---|
| Failover Servers | Backup servers mirror real-time data streams and analytics processes. |
| Automated Backups | Daily and weekly snapshots protect all configurations and historical data. |
| Local Data Caching | SQLite buffers data on each PI during network failures. |
| Self-Healing Services | Scripts auto-restart critical services like MQTT or Kafka in case of crash. |
| Versioned Deployments | Updates are rolled out gradually — ensuring no downtime. |
| Edge Continuity Mode | PI continues to balance load locally even when disconnected from the control center. |
🛠️ The grid never sleeps — neither does its maintenance.
To ensure long-term traceability and analytics potential, the system retains and manages data efficiently:
- Live Operational Data: Stored in TimescaleDB for 1 year.
- Archived Historical Data: Moved to Data Lake (HDFS/NFS) after 12 months.
- Analytical Summaries: Compressed and retained indefinitely for policy insights.
This approach ensures cost-efficiency, scalability, and analytical depth — even over decades of operation.
Reliability isn’t achieved by redundancy alone — it’s designed into every process.
Our software stack is modular, fully open-source, and optimized for scalability and security.
Every layer — from the field edge to the control dashboard — is engineered for real-time performance and government-level reliability.
| Technology | Role |
|---|---|
| Apache Kafka | Message broker handling large-scale data ingestion from PIs. |
| Apache Storm | Real-time stream processing for instant overload and imbalance detection. |
| Apache Druid | Batch analytics for long-term performance trends and capacity planning. |
| TimescaleDB + PostgreSQL | Hybrid database for time-series and configuration management. |
| NGINX + FastAPI (Python) | Backend API and gateway layer for secure system communication. |
| Technology | Role |
|---|---|
| Raspberry Pi OS (Linux) | Local processing platform for edge intelligence. |
| Python Edge Scripts | Executes real-time phase balancing and data filtering. |
| Paho MQTT Client | Publishes encrypted readings to central servers. |
| SQLite Buffer | Local data caching for resilience during outages. |
| Technology | Role |
|---|---|
| React.js | Frontend framework for responsive, modular UI. |
| Grafana | Real-time visualization and system monitoring. |
| Apache Superset | Advanced data analytics and reporting tool. |
| Prometheus | Collects metrics from all running services. |
| Technology | Role |
|---|---|
| TLS 1.3 + AES-256 | Encryption for data in transit and at rest. |
| JWT + RBAC | Authentication and role-based control of system access. |
| VPN + Firewalls | Secure, isolated communication network. |
| Audit Logging System | Tracks all user actions and system events. |
| Protocol | Use |
|---|---|
| MQTT (over TLS) | Lightweight, secure IoT communication between PI and control center. |
| RS-485 (Modbus RTU) | Wired connection between meters and PIs in dense areas. |
| LoRa | Long-range, low-power connectivity in rural networks. |
| HTTP / HTTPS | Secure REST communication for dashboards and APIs. |
| Tool | Function |
|---|---|
| Prometheus + Grafana | Real-time metrics and health visualization. |
| ELK / EFK Stack | Centralized log collection and analysis. |
| Docker | Containerized deployment for scalability and isolation. |
| Git & GitHub | Version control and collaborative development. |
| Attribute | Description |
|---|---|
| Modular Design | Each layer (Hardware, Network, Software) operates independently yet integrates seamlessly. |
| Scalable Infrastructure | Kafka + TimescaleDB handle thousands of parallel streams with low latency. |
| Edge Intelligence | Local PI computation reduces bandwidth and improves reaction time. |
| Secure by Design | Encryption, authentication, and auditing at every level. |
| Interoperable | Compatible with KSEB’s existing SCADA and Smart Grid infrastructure. |
| Cost-Efficient | Fully open-source and on-premise deployment — no cloud cost burden. |
An end-to-end software ecosystem — open, secure, and built for scale.
Our system doesn’t just solve a technical challenge — it transforms how Kerala manages power distribution.
By embedding intelligence into every transformer and home node, the Smart Phase Balancing System delivers measurable operational, economic, and environmental impact.
| Impact Area | Description | Estimated Improvement |
|---|---|---|
| Load Balancing Efficiency | Dynamic switching reduces transformer phase imbalance. | Up to 90% reduction in imbalance duration. |
| Energy Loss Reduction | Reduces technical losses caused by overloading and uneven distribution. | 5–8% decrease in total losses. |
| Transformer Longevity | Balanced loads and predictive alerts reduce wear and stress. | 20–25% longer equipment life. |
| Fault Detection Time | Real-time monitoring detects anomalies instantly. | From hours → seconds. |
| Connectivity Coverage | Hybrid IoT model ensures rural inclusion without heavy infrastructure. | 100% transformer coverage. |
Each percentage saved is power returned to Kerala’s people.
- Reduced maintenance costs due to predictive analytics and automatic fault identification.
- Low operational cost: Minimal recurring expenditure (₹20–₹50/month per PI via IoT SIMs).
- Zero cloud dependency: Fully on-premise architecture prevents third-party expenses.
- Scalable deployment: Compatible with existing transformers — no need for redesign.
- Efficient workforce utilization: Engineers can monitor multiple transformers remotely.
- Energy savings directly reduce the carbon footprint and energy purchase requirements.
- Improved reliability ensures uninterrupted power for industries, homes, and public services.
- Supports renewable integration (solar rooftop and distributed energy sources).
- Empowers rural communities with modern grid access and stable power supply.
- Aligns with India’s vision for Smart, Green, and Sustainable Infrastructure.
🔋 A smarter grid today builds a cleaner tomorrow.
Our project aligns with multiple United Nations Sustainable Development Goals (SDGs), reinforcing its global relevance and long-term purpose.
| SDG | Goal | Contribution |
|---|---|---|
| SDG 7 | Affordable & Clean Energy | Real-time energy optimization ensures efficient and inclusive power access. |
| SDG 9 | Industry, Innovation & Infrastructure | Strengthens national smart grid infrastructure with scalable IoT and data systems. |
| SDG 11 | Sustainable Cities & Communities | Provides reliable power distribution to support urban and rural sustainability. |
| SDG 12 | Responsible Consumption | Reduces wastage and improves the efficiency of resource utilization. |
| SDG 13 | Climate Action | Minimizes energy losses and contributes to lower greenhouse gas emissions. |
Technology aligned with sustainability — progress with purpose.
Behind every system of intelligence lies a team of dedication.
Our group blends expertise from electronics, computer science, and energy engineering — united by one vision: to make Kerala’s power grid smarter, stronger, and sustainable.
| Role | Name | Responsibility | Email ID |
|---|---|---|---|
| Team Lead | Gokul Ram K | System Architecture, Software Design, Integration, API Development, and Data Visualization | gokulram.k2023@vitstudent.ac.in |
| Hardware Engineer | Varun Krishnan R | Home Node & Transformer Module Design & Integration | varun.krishnan2023@vitstudent.ac.in |
| Networking Specialist | Manju Varshikha S | Home Node Design, IoT Connectivity, and Communication Layer Implementation | manjuvarshikha.s2023@vitstudent.ac.in |
| Software Developer | Raghav Sivakumar | Central Dashboard, API Development, and Data Visualization | raghav.sivakumar2023@vitstudent.ac.in |
| Software Developer | Logeswarar G | Data Security, Encryption, and Real-Time Monitoring Modules | logeswarar.2023@vitstudent.ac.in |
| Security & Data Analyst | Karthikeyan Arun | Data Validation, Threat Analysis, and System Documentation | Karthikeyan.arun2023@vitstudent.ac.in |
💡 Innovation is teamwork — and our team powers innovation.
Organization: Smart India Hackathon (Government of Kerala)
Institution: Vellore Institute of Technology, Chennai
Email: gokulram.k2023@vitstudent.ac.in
Location: Chennai, India
Our Demo Work in VIT - Round 2 (Competed Against 850+ teams and made it to TOP 30 after 2 internal rounds!! Internal Round Winners!!)
Building the backbone of a sustainable energy future — one transformer at a time.