Silicon Glades: Edge AI & Industrial Vision with NVIDIA Jetson

Overall Purpose: The purpose of this project is to provide students with hands-on experience deploying hardware-accelerated computer vision and real-time machine learning models at the physical edge using industrial-standard NVIDIA Jetson platforms. Need Addressed: Modern supply chain logistics, smart warehousing, and advanced manufacturing increasingly rely on autonomous visual inspection and edge-based intelligence. While students currently master networking, Linux administration, and foundational microcomputing, they require exposure to dedicated neural processing units (NPUs/GPUs) capable of processing real-time object detection, defect recognition, and automated routing pipelines locally. This project aligns to FL DOE's Computer Systems & Information Technology framework (Program Y100200, CIP 0511090107), specifically Occupational Completion Point A, Computer Systems Technician (CTS0082, 300 hours), which requires students to demonstrate proficiency with computer hardware, troubleshooting, and underlying principles of technology. Deploying and benchmarking vision models on Jetson edge hardware extends this OCP's hardware-proficiency standard into applied edge-AI and computer vision — an emerging-technology skill set not yet codified in the current framework but directly requested by regional logistics and agribusiness employers. Connection to School & District Priorities: Directly aligns with Collier County Public Schools’ strategic focus on workforce development, STEM innovation, and high-wage career placement by closing the skills gap in local automated logistics, agribusiness technology, and technical infrastructure.

Active Student Participation: Working in collaborative technical teams, students will configure NVIDIA Jetson Developer Kits as dedicated "edge nodes." Students will train lightweight vision models, deploy them directly onto the Jetson hardware, and connect high-speed vision cameras to monitor, identify, and sort physical items in real time. Meeting Student & Program Needs: This grant bridges the gap between software algorithms and physical industrial equipment. By integrating these units with our existing Logistics Skill Boss sortation system and previously funded autonomous robotic rovers, students move beyond simulated code to solve real-world industry automation challenges. Instructional Strategies: Project-Based Learning (PBL): Simulated industrial scenarios where teams must design an end-to-end automated quality check station. Collaborative Lab Practicums: Multi-disciplinary setups where edge nodes trigger external robotic actuators upon defect detection. Troubleshooting & Optimization: Benchmarking frame rates, latency, and power consumption across different model quantization layers (INT8/FP16). To maximize hands-on access, students will work in small technical teams that rotate through the three edge-node stations in structured lab blocks — one team actively training and deploying a vision model, while others benchmark previously deployed models, tune quantization settings, or prepare datasets, ensuring every team gets dedicated hardware time each session.

Opening Doors to High-Wage Careers & Generational Mobility: At Immokalee Technical College, our adult learners come from diverse walks of life—many working second jobs, caring for families, or striving to become the first in their households to break into the modern technology economy. The cutting-edge NVIDIA Jetson edge systems will serve as the technical backbone of our new iTECH AI Innovation Lab, placing real-world, industry-standard equipment directly into the hands of students who rarely have access to high-tech resources outside our campus walls. By working with the exact computational tools powering modern logistics, agribusiness, and advanced technology, our students are not just learning theory from a textbook—they are building the skills, confidence, and tangible portfolio projects that unlock life-changing, high-wage careers right here in Southwest Florida. A Campus-Wide Catalyst for Cross-Industry Collaboration: While housed in our IT and Advanced Technical programs, the AI Innovation Lab is designed as an interdisciplinary, campus-wide initiative. Modern automation does not exist in an IT silo; it touches every field our college trains for: Manufacturing: Students connect our Amitrol Skill Boss trainer sorting system with intelligent cameras to learn automated package handling and inventory flow. Agriculture & Environmental Tech: Learners explore how to support local farms in the tech they are already using for instance: using computer vision to spot crop distress, sort produce, and conserve water. Accounting, Business Operations & Inventory Control: Modern enterprises no longer track assets with clipboards or manual spreadsheets. By pairing vision models with real-time tracking, students bridge IT and accounting—learning how automated edge systems log physical inventory into digital ledgers, calculate real-time cost-of-goods, track depreciation, and flag stock discrepancies instantly to prevent financial loss. Students across different certificate programs will collaborate in our lab, discovering how disparate industries interconnect and solving complex problems together just like cross-functional teams in the professional workforce. Success will be measured through: (1) a completed end-to-end vision pipeline (trained model + live camera deployment) per team, scored against a rubric for detection accuracy and integration with the sortation system; (2) benchmarked performance metrics — frame rate, inference latency, and power draw — logged across model quantization settings (INT8/FP16) to demonstrate measurable optimization gains; (3) a final demonstration in which teams present a working automated quality-check station to peers and instructors; and (4) a portfolio artifact (code, benchmark data, and demo video) each student can use in technical job interviews.

The requested funds will acquire the core physical compute and optical infrastructure needed to establish the iTECH AI Innovation Lab at Immokalee Technical College. While foundational networking and microcomputing are taught in our classrooms, our adult learners currently lack access to dedicated Edge AI hardware with hardware-accelerated tensor cores. Funding will directly provide student workstations with NVIDIA Jetson edge compute units, high-speed camera sensors, fast solid-state storage, and safe bench-mounting breakout equipment. Rather than relying on cloud simulations—which carry recurring subscription costs and hide the physical realities of edge networking—these one-time grant funds place real, industry-standard embedded computing hardware directly into students' hands. Industrial M.2 NVMe SSD (512GB High-Endurance) – $55: This drive provides dedicated, high-endurance local storage for the Jetson Orin Nano edge workstation, which will continuously read and write large volumes of vehicle telemetry, video feed data, and trained model checkpoints throughout repeated student lab sessions. A high-endurance industrial-grade SSD is required (rather than a standard consumer drive) because the constant read/write cycles from real-time inference logging and iterative model training would quickly degrade standard storage, risking data loss and downtime during instructional use. The three Jetson units requested here are dedicated exclusively to powering the camera-based computer vision pipeline and integrating with our existing Logistics Skill Boss sortation system and autonomous rovers, distinct from the edge-compute hardware requested in our companion DeepRacer and sensor-audit proposals.

1980

3 - NVIDIA Jetson Orin Nano Developer Kit (8GB)

1497

3 - IMX477 High-Quality 12.3MP Camera Modules + Lens Kits

267

3 - Industrial NVMe SSDs (512GB M.2 2280)

165

3 - Enclosed Mounting & Power Breakout Kits

51