Many customers—individuals, small merchants, micro-entrepreneurs and local logistics players—need short-term, trustworthy tracking but cannot afford to buy expensive devices or manage them long-term. Examples:
A parent wants to monitor a child’s travel to/from school without buying a device for occasional use.
A home-cook startup outsources deliveries and needs to verify package hand-off and route integrity.
Small movers/packers need temporary tracking during transit.
TrackSwap solves this with a rented device + subscription model (deposit + short-term fee). Devices are returned for deposit refund and reused. The core differentiator is an AI agent that does the monitoring and takes actions for the user (phone calls, SMS, escalation to backup contacts, automated rescheduling with delivery partners, or start a remote recovery workflow) so the user doesn’t need to constantly watch dashboards.
Key user flows:
Merchant or user requests a tracker for a period (hour/day/month) via web/app.
Device shipped or collected with refundable deposit. Device registers with TrackSwap via AWS IoT Core.
Device transmits location & health telemetry to AWS backend.
AI Agent (Bedrock + agent orchestration) monitors geofences, routes, battery and tamper events.
On anomalies (geofence breach, tamper, low battery, unexpected stop), the agent follows a policy: attempt automated call to the user → call backup contact → notify delivery partner → raise ticket to recovery team. Billing is usage-based; unused deposit returned after device return and inspection. Business benefits: Low entry cost for tracking (rental model). Trust & accountability for micro-entrepreneurs and customers. Reduced asset loss and better delivery efficiency for partner platforms (marketplaces can white‑label TaaS). Increased device utilization and circular lifecycle. High-level architecture (text): Devices → Cellular (LTE‑M) send MQTT/HTTPS telemetry to AWS IoT Core. IoT Rules route telemetry to Lambda and IoT Analytics + write raw to S3. Lambda enriches event, writes device state to DynamoDB and emits events to EventBridge/SQS. Bedrock-backed Agent (invoked via Lambda) evaluates policies (geofence via Amazon Location Service, tamper, dwell time, battery) and decides actions. Actions: send push/SMS via SNS/Pinpoint, place voice calls using Amazon Connect, or call partner APIs via API Gateway. Billing microservice (Lambda + DynamoDB) calculates usage and triggers payment events. IoT Device Management manages fleet, provisioning, OTA updates. Monitoring/logging via CloudWatch and user-facing dashboards served through API Gateway + static UI in S3/CloudFront Integrated Amazon Bedrock agent to handle escalation flows and natural-language phone/SMS interactions. Added automated geofence policies and test harness for device simulation. Implemented pay-as-you-use billing microservice and deposit handling in DynamoDB AWS Services AWS IoT Core (device connectivity & device shadow) AWS IoT Device Management (fleet onboarding, OTA updates) AWS Lambda (stateless compute for event handlers) Amazon API Gateway (REST API endpoints) Amazon Bedrock (LLM / intelligent agent brain) Amazon SageMaker (optional ML models for route anomaly / dwell-time predictions) Amazon SNS & Amazon Pinpoint (SMS/Push notifications) Amazon Connect (automated voice calls / IVR for escalation) Amazon SQS & Amazon EventBridge (event orchestration & retries) Amazon DynamoDB (device registry, sessions, billing records) Amazon S3 (telemetry archive, device logs, uploaded images) Amazon Location Service (geofencing, mapping) AWS CloudWatch (monitoring & alarms) Device / Hardware & Connectivity Compact GPS tracker with LTE-M / NB‑IoT + BLE (reusable industrial tracker) eSIM for cellular connectivity (or customer-supplied SIM) Lightweight tamper switch and accelerometer for motion/tamper detection
Built With
- amazon-dynamodb
- aws-iot-core
- bedrock
- lambda
- s3
Log in or sign up for Devpost to join the conversation.