AI BRIEFCASE delivers teams full-stack AIML capability, even in disconnected situations.
No compromise.
AI Briefcase
Features
Directly ingest, store and analyse data in edge situations
Rely on ready-to-go robust software & rugged hardware
Work with confidence in a secure, disconnected environment
All-in-one IOT/AIML Platform
Ingest Module
Ingest or push data from any external applications, IoT devices or formats using Rest API
Input options include MQTT, file systems, relational DBs, CSV, JSON
Data-Lake Module
Securely store data within the device for discovery, analytics and machine learning
AI/ML Workbench Module
Develop, Train, Deploy Machine learning models within the edge briefcase
Full-Stack Workbench Module
Develop, Test, Deploy Applications that use ML Models or provide analytics on data from the data-lake
Analytics Module
Dashboard & Reports UI
REST API for accessing inference from deployed ML Models
Benefits
Site Specific
Situations where environmental conditions can affect technology reliability and reliable data capture
Mining Engineering, Tropics Meteorology, Outback Remote Testing
Mining Engineering, Tropics Meteorology, Outback Remote Testing
Speed Sensitive
Situations where connectivity issues and time to data processing and analysis are priority factors
Maritime Data, Remote Communities, Research Facilities
Maritime Data, Remote Communities, Research Facilities
Strength & Security
Situations where requirements mandate isolating or disconnecting the platform to create a secure private environment
Defence Projects, Critical Infrastructure Assets, Utilities
Defence Projects, Critical Infrastructure Assets, Utilities
Hardware Specifications
Compute and memory
| CPU | Up to 104 vCPUs (available in configurations of 52 or 104 vCPUs) |
| RAM | 512 GB RAM (Up to 416 GB RAM – Customer usable) |
| GPU | nVidia V100 (available in Compute Optimized with GPU configuration – only offered with 52 vCPU) |
| Storage | SSD storage capacity 28 TB NVMe SSD or 42 TB HDD (39.5 TB usable) |
Compute and memory
| CPU | Up to 104 vCPUs (available in configurations of 52 or 104 vCPUs) |
| RAM | 512 GB RAM (Up to 416 GB RAM – Customer usable) |
| GPU | nVidia V100 (available in Compute Optimized with GPU configuration – only offered with 52 vCPU) |
| Storage | SSD storage capacity 28 TB NVMe SSD or 42 TB HDD (39.5 TB usable) |
Dimensions and weight
| Weight | 49.7 pounds (22.45 Kg) |
| Height | 15.5 inches (394 mm) |
| Width | 10.6 inches (265 mm) |
| Length | 28.3 inches (718 mm) |
Environment
| Vibration | Non-operational use equivalent to ASTM D4169 Truck level I 0.73 GRMS |
| Shock | Operational use equivalent to 70G (MIL-S-901), Non-operational use equivalent to 50G (ISTA-3A) |
| Altitude | Operational use equivalent to 0–3,000 meters (0–10,000 feet), Non-operational use equivalent to 0–12,000 meters |
| Temperature range | 0–45°C (operational) |
Achieve your decision advantage with AI Briefcase
Software Specifications
| Component | Software |
| Virtual Servers | Ubuntu 22.04 LTS |
| Container Management | OPortainer 2.18.x |
| Containers | Docker Swarm + Docker |
| Version Control | GitLab CE |
| ML Workbench | VSCode + Jupyter |
| Full-Stack Workbench | VSCode + Boto3 |
| REST API Endpoint | FastAPI |
| Added Security Layer | NGINX Reverse proxy |
| Other tools | Docker Registry |
Online/Offline Flexibility
Online mode – Connected to the cloud/internet
Offline mode – Totally disconnected from the cloud/internet in high security zones such as defence environment or at remote locations with limited or no internet connectivity (the outback or at sea)
ML OPS AT THE EDGE
- Create, train, test and deploy algorithms at the edge
- Monitor and manage applications
- Version Control for deployments and rollbacks
- Industry-leading GitLab CE Instance
- Replicate code to other briefcases (primary backup) or cloud, such as S3 (secondary backup)
- Local Docker registry with commonly used images, ability to add any image(s)
