> For the complete documentation index, see [llms.txt](https://wattsee.gitbook.io/wattsee-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://wattsee.gitbook.io/wattsee-docs/our-mvp.md).

# Our MVP

## Core Feature

1. **Realtime Monitoring (basic)**
   * Track voltage, current, and power.
   * Log all charging data.
2. **AI-based Anomaly Detection**
   * AI uses baseline charging profiles (onboarding).
   * Detects early anomalies such as:
     * Charging power consistently lower than expected.
     * Abnormal voltage fluctuations.
     * Current spikes not matching previous charging patterns.
3. **AI Session Analysis**
   * After each charging session, AI reviews all data:
     * Assess overall safety level.
     * Compare against past sessions.
     * Provide insights like: *“Session stable”*, *“Potential risk detected”*, *“Check battery/charger”*.
4. **Self-Learning from Data**
   * Each time the user confirms *“real issue”* or *“false alarm”*, AI stores feedback.
   * Gradually builds a personalized **charging safety model**.

***

## User Experience MVP

* **Realtime Dashboard**: show V, A, W + AI status (*Safe / Risk Detected*).
* **AI Alerts**: push notification when anomalies occur.
* **AI Reports**: post-session summaries highlighting risks and trends.

## Technical Scope (updated)

* **Hardware**: use third-party smart plugs / energy meters; sample and collect telemetry every **5 seconds** (V, A, P, timestamp).
* **Backend + AI service**: ingest 5s telemetry, run rule-based checks (spike, thresholds) and AI analysis (early-window + full-session), store sessions and user feedback for continuous learning. Provide API endpoints for realtime status, alerts, and session reports.
* **Mobile app / Web**: realtime dashboard (V/A/W every 5s), push notifications for AI alerts, post-session report view, and a simple onboarding flow to capture charging profile metadata (battery capacity, charger rating, etc.).

## Video Demo

{% embed url="<https://www.canva.com/design/DAGzhRXcgPc/wBWnpdx6R-zzxV7duMuE0A/watch?utlId=h659fabc886&utm_campaign=designshare&utm_content=DAGzhRXcgPc&utm_medium=link2&utm_source=uniquelinks>" %}

## What issues Wattsee AI can detect?

#### 1. Electrical Issues

* **Abnormal voltage**:
  * Large fluctuations >±10% from normal.
  * Undervoltage → slow charging, overheating wires.
  * Overvoltage → risk of damaging the BMS or charger.
* **Abnormal current**:
  * Sudden current spikes (possible short circuit or fault).
  * Unstable or oscillating current during charging.
  * Abnormally low current (vehicle not charging even though power is supplied).
* **Abnormal power**:
  * Persistently low charging power (vehicle/battery not accepting charge).
  * Strong fluctuations in power (sign of poor socket contact or weak wiring).
  * Irregular charging pattern that doesn’t follow the normal CC–CV curve.

#### 2. Connector & Load Issues

* **Overload risk**: total household power exceeding safe threshold.
* **Overheating indicators**: inferred from high current + strong fluctuations → potential melting or burning of connectors/outlets.
* **Loose connection**: charging interrupted multiple times within a session.

#### 3. Charging Behavior Issues

* **Unexpected interruptions**: charging stops abruptly without vehicle command.
* **Unusual session duration**: charging takes too long without significant battery gain.
* **Profile deviation**: compare with historical charging profiles → detect anomalies (e.g., same EV, same battery, but very different current/power).