Fleet Tracking & Teleoperation
Edgenesis partnered with SS Rover to develop a comprehensive teleoperation solution. The solution successfully integrates multiple types of robotic machinery including ATVs, Robot Arms, Robot Dogs, and various mobile robots, enabling unified remote control and monitoring capabilities through a standardized TeleOps Kit. Through this collaboration, SS Rover is able to expand their operational capabilities, offer remote operation services to various industries, automate operations through AI, and standardize their machine integration process for rapid scaling.
The solution leverages edge computing and cloud technologies following a hybrid architecture. Data undergoes transformation and filtering at the edge through FaaS processing, while the cloud infrastructure manages video streaming, monitoring, teleoperation, and digital twin services. The solution ensures stable connectivity through network redundancy and secure VPN connections.
Edgenesis delivered the following to SS Rover:
- TeleOps Kit: A standardized hardware and sensor package that transforms any compatible machinery into a remotely operable SS Rover-enabled device
- Teleoperation (TeleOps): Control multiple types of machines remotely through a unified interface with stable video streaming and MQTT communication
- Device Integration: Seamless integration of diverse machinery and devices with ROS2, Modbus, S7 and other propretery protocols through Shifu
- Edge Computing: Distributed computing capabilities through Edge Kubernetes clusters, ensuring efficient local data processing and reduced latency
- Edge AI: Enable real-time AI inference at the edge for vision-based and language-based tasks, bringing intelligence directly to the devices
- Cloud AI Training: Infrastructure for training and updating AI models in the cloud, with automated deployment to edge clusters
- Standardized Deployment: Unified approach to machinery integration enabling consistent deployment across different device types and geo-locations
- Digital Twin: Unity-based virtual environment providing real-time visualization of machine positions and orientations
- Network Resilience: Balanced connectivity across multiple network types (4G/5G, Starlink, Ethernet) ensuring stable operations
- Security and RBAC: Providing fine-grained access control of all machines, preventing security breach via end to end encryption and VPN tunnel
Challenges
- System Architecture: Designing a scalable architecture that can handle multiple types of machines, high-bandwidth video streams, and real-time control while maintaining low latency and high reliability
- Network Design: Implementing a reliable network stack that can handle multiple connection types (4G/5G, Starlink, Ethernet), automatic failover, and maintain stable connections for critical operations
- Protocol Integration: Building protocol adapters in Shifu to handle diverse communication protocols (ROS2, Modbus, S7, proprietary protocols) while ensuring consistent data transformation and control interfaces
- Edge Orchestration: Managing distributed Kubernetes clusters at the edge with limited computing resources while ensuring AI inference, video streaming, and control operations run smoothly
- Security Implementation: Implementing end-to-end encryption and RBAC without compromising real-time performance, while ensuring secure remote access across different network environments
- Deployment Strategy: Creating a standardized deployment process that works consistently across different environments and geographical locations, while maintaining system performance
Edgenesis Solution
Standardized Edge Infrastructure
A Kubernetes-based system ensures adaptability to diverse devices and tele-operation use cases, promoting both scalability, high availability and security
Centralized Application Management
Effortlessly deploy and manage applications across edge and cloud clusters from a single interface. Seamlessly roll out updates and, if needed, revert changes with a single click
Scalable Cloud Storage with HA
A versatile cloud storage solution featuring relational databases, time-series databases, and object storage accommodates various data types. On-demand expansion capabilities address future growth needs
Data Processing at the Edge
Optimize bandwidth usage and facilitate real-time decision-making by processing data at the edge. This allows for video compression and filtering of unnecessary data
Core Services
Extensible and scalable design of core fleet management platform services lays the foundation for future enhancements
Seamless Thingsboard Integration
Visualize device data directly in Thingsboard for streamlined monitoring and insights
Results
A Thingsboard based UI platform that enables real-time data insight and control.
