TRUSTED AUTONOMY PORTAL -2.pdf.crdownload

REQUEST FOR BIDS/PROPOSALS COVERSHEET

THE UNIVERSITY OF SOUTHERN MISSISSIPPI

Procurement and Contract Services

118 College Drive #5003, Hattiesburg, Mississippi 39406-0001

Date: February 24, 2025

BID No. 25-36

THE UNIVERSITY OF SOUTHERN MISSISSIPPI is considering the purchase of the following item(s). We ask that you submit your bid and retain one copy for your files. Right is reserved to accept or reject any part of your bid. Your quotation will be given consideration if received in Bond Hall, Room 214 on or before:

2:00 p.m. CT TERMS - Bidder should state terms of sale. Our terms are 2% ten days, net 45 days.

These terms will apply per Mississippi law. March 13, 2025 AWARDING CONTRACT - Cash terms will not be used as a basis for awarding contracts; however, the University will accept cash discounts when earned. Buyer: Amber Floyd

NOTE: If you cannot quote on the exact material shown, please indicate any exception giving brand name and complete specifications of any alternate. If additional space is required, use a separate sheet or letter of transmittal.

ITEM

QUANTITY

DESCRIPTION

BID 25-36 Trusted Autonomy Portal

RFx # 3160007235

UNIT PRICE

TOTAL NET PRICE

PROPOSAL MUST BE RETURNED TO THE UNIVERSITY IN

ACCORDANCE WITH THE SPECIFICATIONS. RFP NUMBER AND

DATE OF BID OPENING MUST BE SHOWN ON THE OUTSIDE OF

THE ENVELOPE IF USING THAT METHOD.

We quote you as above - F.O.B. The University of Southern Mississippi.

Shipment can be made in days from receipt of order. DATE ______________ Return quotation to Procurement Services at above address.

AA/EOE/ADAI Signature Required__________________________________________

THIS IS NOT

AN ORDER

Name: ____________________________________

Company: _________________________________

Address: __________________________________

City/State/Zip: ______________________________

AA/EOE/ADAI

Specifications for an Ocean Enterprise Trusted Autonomy Portal (Hardware and Software)

The University of Southern Mississippi

Prepared by Steve Stanic and Landry Bernard

Feb 24/2025

Specifications for an Ocean Enterprise Trusted Autonomy Portal (Hardware and Software)

I. Purpose

This RFP document outlines the specifications for a Trusted Autonomy Portal (TAP). The portal system shall utilize state-of-the-art technologies, including AI, machine learning, simulation engines, existing maritime systems, and models. The system will include both hardware (high-performance laptops/desktops, IoT devices) and software (AI models, simulation tools, data integration systems) components. The scope of the project includes the design, development, testing, and deployment of the TAP.

II. Background This document defines the requirements for the development and deployment of an Ocean Enterprises Trusted Autonomy Portal. It shall include AI-driven simulations and be deployable on high-end laptops or desktops for field operations. The TAP is intended for maritime research, autonomous vessel testing, and decision-making applications, adhering to strict security standards and delivering robust performance in varied environments. The TAP should provide scalable, secure, and efficient solutions for handling maritime autonomy scenarios, integrating advanced oceanographic models. The system shall comply with all relevant maritime data standards for vessel tracking, navigation, simulation, DoD standards for security, encryption, and data management.

III. Specific Functional Specifications and Requirements

Integration with Existing Models

- The TAP shall provide AI-driven simulations of maritime scenarios. The AI shall also predict vessel behaviors, optimize collision avoidance strategies, and simulate environmental interactions.

Integration with Existing Models

- Bellhop 2D/3D Models: Simulate sound propagation and underwater acoustic scenarios.

- NGOFS2/HYCOM Models: Simulate Gulf of Mexico conditions, including hydrodynamic, current, and wave models.

- USM’s Hi-Res Coastal Models: For high-resolution coastal and port simulations.

- AI Models: Shall support ONNX-based models for performance optimization on the target hardware.

- Simulation Tools: These tools shall support integration with tools for real-time 3D vessel behavior modeling and environmental simulation.

Data Ingestion and Processing

- The system shall be capable of ingesting data from maritime sensors, autonomous vessels, IoT devices, and environmental monitoring systems.

- Data shall include but is not limited to vessel position, speed, heading, and telemetry.

- Environmental data such as water temperature, wind speed, wave height, and obstacles shall also be available.

Ingestion Protocols

- The system shall support data ingestion via MQTT, Kafka, WebSocket, and REST APIs.

Data Handling

- The system shall normalize and clean incoming data streams to ensure consistency across all simulation scenarios.

Edge Computing Deployment

- The TAP shall support deployment on high-performance laptops/desktops allowing for field-based use. Some features may be cloud based.

- Laptop Specifications: High-end laptops with dedicated GPUs, supporting both Windows and Linux shall be provided.

- On-Device AI Inference: Only ONNX and TinyML shall be used to optimize AI model inference on devices with limited resources.

- Scalability: The system shall be capable of scaling from local deployment on a single laptop to a cloud-based environment for more extensive simulations.

Secure Data Management

- Ensure that all data within the TAP is securely managed, stored, and transmitted.

This shall include the following:

- Encryption: End-to-end encryption of all data streams and stored data.

SME Tools for AI Model Creation

- The vendor shall provide user-friendly tools for Subject Matter Experts (SMEs) to create, update, and deploy AI models within the simulation environment.

This includes the following:

- Jupyter Notebook Integration: Allow SMEs to interact with AI models and datasets using Jupyter Notebooks.

- Python Libraries: Provide access to common libraries such as TensorFlow, PyTorch, SciPy, and Xarray for model development and data analysis.

- Model Deployment: SMEs shall be able to deploy their AI models to the simulation environment without requiring in-depth technical knowledge.

IV. Specific Non-Functional Specifications and Requirements

Scalability

- Vertical Scaling: The system shall scale with the complexity of the simulations, allowing for additional compute resources to be added as needed.

- Horizontal Scaling: The system shall support the addition of multiple nodes for cloud-based deployment when scaling beyond single machine processing power.

Performance

- Latency: The system shall process real-time data with a latency of no more than 1000ms.

- Throughput: There shall be Support for at least 50,000 data points per second during ingestion and processing.

- Simulation Speed: The system shall run simulations efficiently, handling up to 100 vessels simultaneously in real-time.

Security

- Compliance: The system shall comply with DoD cybersecurity standards, including encryption, secure access control, and data integrity measures.

- Audit Logging: A detailed audit log of all actions performed within the system shall be provided, including data access, model changes, and simulation outcomes.

Reliability and Availability

- Uptime: The system shall maintain at least 99.9% uptime

- Disaster Recovery: The system shall have disaster recovery mechanisms in place, with automatic backups and recovery within 4 hours of an incident.

Usability

- User Interface: The system shall provide an intuitive user interface for configuring simulations, monitoring data streams and adjusting AI models.

- SME-Friendly Tools: Shall provide non-technical users with the ability to interact with AI models and simulations without requiring coding expertise.

V. Specific Specifications for Each Laptop Computer The following minimum specifications that each Laptop Computer shall have to support USM’s project objectives.

VI. Specific Specifications for Desktop Computers The following minimum specifications that each Desktop Computer shall have to support USM’s project objectives.

VII. Deliverables

The following is a list of specific deliverables but not be limited to only the following:

1. A complete architectural design document detailing how the system integrates all components shall be delivered (AI, blockchain, simulation tools, etc.).

2. High-performance laptops, desktops, and IoT devices shall be delivered.

3. Delivery of the TAP software shall include installation and deployment instructions.

4. Pre-trained AI models for collision avoidance and vessel behavior with documentation for model retraining shall be delivered.

5. A suite of tools and user documentation for SMEs to interact with the simulation environment and AI models shall be delivered.

6. Performance, scalability, security, and usability testing reports shall be delivered.

Laptop Specifications CPU The CPU shall be an Intel i9 or AMD Ryzen 9 equivalent or better GPU The GPU Shall be NVIDIA RTX 3080 or equivalent, capable of handling

AI inference workloads RAM Shall be at least 32GB Storage At least 1TB SSD Operating System Shall Support for both Windows 10/11 and macOS (M3 Pro/Max) Data Sensors Must support integration with IoT sensors for real-time data collection, including GPS, environmental monitoring, and autonomous vessel telemetry

Connectivity The system shall support multiple communication protocols for IoT devices, including 5G, Wi-Fi 6, and LoRa

Desktop Specifications CPU The CPU shall be AMD Ryzen Threadripper 5995WX or Intel i9-13900K

GPU The GPU Shall be NVIDIA RTX 4090 or NVIDIA A100 (for AI/ML workloads) RAM Shall be at least 128 GB or 256GB DDR5 Storage At least 2 TB NVMe SSD (Primary) + 4TB NVMe SSD (Secondary) Operating System The operating system shall be Windows 11 Pro / Ubuntu 22.04 LTS Cooling The cooling system shall be liquid or high-performance air cooling

PSU 1600W, 80 PLUS Titanium Motherboard ASUS ROG Zenith II Extreme Alpha (Threadripper) or ASUS ROG

Maximus Z690 Hero (Intel)

7. The delivery arrangements and transportation costs shall be the vendor’s responsibility.

8. These Tap systems, software packages, and support equipment shall be delivered to

USM’s Marine Research Center, 1030 30th Ave Gulfport, MS 39501.

VIII. Training The vendor shall provide training on the operation, software, maintenance, troubleshooting, and integration of all systems. This training shall take place at the USM facility located in Gulfport MS.

IX. Proof of Performance

The vendor shall provide a detailed technical approach to meet the functional and non-functional requirements outlined in this document. A summary of relevant experience in maritime autonomy, AI, and IoT integration and a detailed project timeline with milestones and estimated delivery dates shall be provided. The vendor needs to provide references where their AI systems in a similar configuration as specified above have operated successfully within the last 2 years. The above requested information will assist USM in determining the bidder’s capability of meeting these requirements.

X. Warranty Services

At a minimum, the Contractor shall provide software/hardware warranty support for one year from acceptance. Longer warranty periods are preferred. The Vendor shall agree to repair, adjust, and/or replace (as determined by the University to be in its best interest) any defective materials at the Vendor and/or manufacturers’ sole cost. The University will incur no costs for service or replacement of materials during the warranty period.

The Vendor will be the sole point of contact for warranty issues.