Equipment Checklist 0 / 19 checked
1

Equipment List (Bill of Materials)

Check items as you gather them. Progress is saved automatically. Click any item row to toggle.

Compute & Control
Cameras (×5 GoPro Rig)
Mounting & Power
Athlete Identification
Estimated Kit Cost Summary
ItemQtyUnit CostTotal
NVIDIA Jetson Orin Nano Super 8GB1$249$249
GoPro HERO 13 Black5$399$1,995
GoPro Media Mod5$79$395
Elgato Cam Link 4K5$129$645
Heavy-duty tripods6~$60~$360
WiFi 6 Router1~$120~$120
Powered USB 3.0 Hub1~$50~$50
Accessories (cables, sandbags, power banks, etc.)~$179
Full 6-Camera Rig Total~$3,993
1-Camera Pilot (Jetson + 1 cam setup)~$1,100

2

Phase 1 — Pre-Event Setup

Phase 1 · 24–48 hrs before

Complete these steps at home or office before the event day. Skipping pre-event prep is the #1 cause of field delays.

1
Charge Everything
Fully charge all hardware before leaving for the event.
  • GoPro batteries — 2 per camera recommended (10 total)
  • iPad Pro and iPhone 15 Pro to 100%
  • Jetson Orin via wall adapter (do not rely on USB-C alone)
  • Both 10,000 mAh power banks to full
2
Configure GoPros
Apply these settings to every HERO 13 Black before the event:
  • Video: 1080p Superview, 240fps (use 120fps for the 4K overview camera)
  • Stabilization: OFF — HyperSmooth causes latency that breaks frame timing
  • Protune: ON — enables manual ISO control and flat color profile
  • White Balance: Auto (or lock to 5500K for consistent outdoor daylight)
  • ISO Limit: 1600 max — prevents motion blur in high-ISO noise reduction
3
Flash Jetson SD Card
Download the Rep 1 Exposure Edge Agent image from app.rep1.io/downloads. Flash to a 64 GB+ microSD card using Balena Etcher (free). Insert into Jetson — first boot takes approximately 4 minutes while the system expands the filesystem and starts services.
4
Pre-Load Athlete Roster
Log into app.rep1.io → Events → select your event → Import Athletes. Upload a CSV (name, phone, position, school) or add manually. Assign ArUco IDs to each athlete in the portal — this links the printed marker number to their profile.
5
Print ArUco Markers
Admin Console → Check-In → Print Markers. Print on cardstock (not regular paper). Laminate each sheet. Number the back of each marker to match the athlete assignment. Minimum print size: 4"×4". Prepare at least 60 for a full event, plus 10 spares.
6
Test Network at Home
Power on the Jetson and router together. On your iPad, open a browser and go to rep1-admin.local (or 192.168.1.100). Verify all 6 camera slots appear in the feed grid — plug in USB webcams if you don't have all GoPros available yet. Each feed should show a green "Live" indicator. Confirm the Admin Console roster matches your imported athletes.
7
Calibrate AI Timing Baseline
Run 3 staff members through the 40-yard course at your practice space. Admin Console → AI Settings → Run Calibration. The system calculates its frame-to-distance ratio using the known 40-yard distance. Calibration takes about 2 minutes to process. Store the calibration profile to the event — it will be applied automatically on event day.

3

Phase 2 — Field Setup

Phase 2 · 90 min before start
Camera Placement Diagram — Top-Down View · 40-Yard Dash Lane
Athlete runs left → right 0 yd 10 yd 20 yd 30 yd 40 yd Check-in iPhone 15 Pro Check-in GoPro 1 Start Line 0 yd GoPro 2 Accel Phase 10 yd GoPro 3 Mid-Sprint 20 yd GoPro 4 Finish Line 40 yd GoPro 5 Overview Elevated Jetson + Router Sideline USB hub cables (dashed = cable runs to Jetson)
iPhone 15 Pro (Check-in ID)
GoPro HERO 13 (Timing camera)
Jetson Orin + Router (Sideline compute)
Cable runs (HDMI→Cam Link→USB)
Camera orientation: Mount all cameras perpendicular to the athlete's running path — not head-on. Perpendicular angles give the AI the cleanest frame-by-frame position data for split-time calculation.
1
Layout Tripods per Diagram
Place tripods at: check-in area, 0 yd (start), 10 yd, 20 yd, 40 yd (finish), and one elevated overview position behind the finish. Set all camera heights to 3–4 ft (waist height) for optimal ArUco marker detection. Confirm all cameras face perpendicular to the lane — not angled toward the athlete's path.
2
Mount GoPros + Media Mods
Slide each Media Mod onto the GoPro (snaps in place from below). Connect: GoPro HDMI out → Cam Link 4K → USB-C cable → USB hub port. Secure every cable connection with gaffer tape — field vibration and wind can loosen HDMI connectors mid-event. Label each Cam Link 1–5 matching its camera position.
3
Power On Jetson
Connect the Jetson power adapter. Wait for the green LED (approximately 90 seconds). SSH into the device or connect an HDMI monitor, then run: sudo systemctl start rep1-agent. The agent will initialize camera detection automatically within 30 seconds.
4
Connect All Devices to Router
Jetson connects via ethernet cable to the router (preferred). iPad and iPhone join WiFi network REP1-FIELD (password: rep1field2026). No internet connection is required — all AI processing runs locally on the Jetson. The router only manages local device communication.
5
Open Admin Console — Verify All 6 Feeds
On iPad, navigate to rep1-admin.local (or 192.168.1.100 if mDNS is unavailable). Verify all 6 camera feeds appear in the grid. Each feed should display a green status dot. If any camera shows gray/offline, see Troubleshooting → "Camera feed not showing."
6
Set Camera Zones
Admin Console → Camera Zones. Assign each of the 6 feed slots to its zone: Start Line (0 yd), 10 yd Split, 20 yd Split, 40 yd Finish, Overview, and Check-in iPhone. Zone assignments tell the AI which split time each camera is responsible for measuring.
7
Set Minimum Confidence Threshold
Admin Console → AI Settings → Min ID Confidence. Set to 75%. Clips where athlete identification confidence falls below 75% will automatically route to the manual review queue rather than auto-approve. This prevents misidentified clips from going to the wrong athlete profile.
8
Run a Full Test Athlete
Walk one staff member through the complete drill as an athlete:
  • Scan face or ArUco marker at iPhone check-in station
  • Attach ArUco marker to back of jersey
  • Walk/jog the 40-yard lane
  • Confirm: clip triggered on at least 3 cameras → clip appears in review queue → approve clip → confirm athlete profile is updated
If anything in this chain fails, resolve it before athletes arrive.

4

Phase 3 — Event Day Operations

Phase 3 · Live Operations
Check-In Station (iPhone)
  • Open Rep 1 Exposure Check-In app on iPhone
  • Scan athlete face or scan their ArUco QR assignment sheet
  • Assign the printed ArUco marker — stick to jersey back, centered between shoulder blades
  • Face-enrolled athletes: no marker required after initial scan
  • Confirm athlete shows "Checked In" in Admin Console → Roster
During Drills
  • System runs automatically — no button presses needed
  • Operator monitors Admin Console → Camera Feeds for offline cameras
  • Review queue: approve clips with ≥80% confidence automatically
  • Flag unusual clips for manual review
  • Athletes and parents are notified via SMS automatically if phone number is in their profile
Post-Event Quick Export
  • Admin Console → Events → Export CSV
  • Exports: timing data, clip list, and scout view stats
  • Send to event organizer immediately after last drill
  • Athletes receive their profile link via SMS within 10 minutes of export
Monitoring Checklist
  • Check Jetson CPU temp every 30 min — alert if above 80°C
  • Verify GoPro battery levels — swap if below 20%
  • Check upload queue doesn't exceed 50 pending clips
  • Confirm router is broadcasting — re-boot if any device drops

5

Phase 4 — Tear Down

Phase 4 · ~30 min
Do not power down until the upload queue is clear. Powering off the Jetson with pending clips can corrupt the upload buffer and cause data loss.
1
Wait for Upload Queue to Clear
Admin Console → Upload Queue. Wait until the pending count reaches 0. On good WiFi, this takes 5–15 minutes after the last drill. Do not begin tear down of any cameras or the Jetson until the queue is empty.
2
Force-Sync if Queue is Stuck
If the upload queue has been stuck at the same count for more than 5 minutes, go to Upload Queue → "Force Sync All". This bypasses the incremental queue and pushes all pending clips directly to Azure Blob Storage. Requires at least minimal internet connectivity from the Jetson.
3
Export Local Backup
Admin Console → Events → Export Full Backup (JSON). This saves a complete event record — timing data, clip metadata, athlete list, and confidence scores — to the Jetson's SD card as a redundancy copy. Copy this file to a USB drive as well if possible.
4
Power Down in Order
Always shut down in this sequence to avoid data loss:
  • GoPros — press and hold power button until off
  • iPad — close Admin Console browser tab, then sleep
  • iPhone — close Check-In app
  • USB hub — unplug from Jetson
  • Jetson — SSH in or use monitor and run: sudo shutdown -h now. Wait for green LED to go off before unplugging power.
  • Router — power off last
5
Coil and Pack Cables
Coil all USB-C and HDMI cables using the over-under technique to prevent kinks and cable memory damage. Store each GoPro in its padded protective case. Wrap Cam Links in their original pouches or small zip bags. Keep all cables labeled for next event setup.
6
Check Battery Status for Recharging
Note any GoPro batteries, power banks, or device batteries below 20% charge. Charge all low batteries before the next event — low GoPro batteries cause automatic framerate drops to 60fps that break timing accuracy.

6

Camera Placement Quick Reference

Standalone reference for printing. Laminate and keep at the setup crew bag.

Camera Positions at a Glance
DevicePositionHeightAngleFrame RateZone
iPhone 15 ProBefore start — check-in area5 ft (face height)Front-facingID / Check-in
GoPro 10 yd — start line3–4 ftPerpendicular240 fpsStart Explosion
GoPro 210 yd mark3–4 ftPerpendicular240 fps10-yd Split
GoPro 320 yd mark3–4 ftPerpendicular240 fps20-yd Split
GoPro 440 yd — finish line3–4 ftPerpendicular240 fps40-yd Finish
GoPro 5~60 yd back — elevated6–8 ftWide angle toward start120 fpsOverview

7

Troubleshooting

Click any issue to expand the cause and fix.

Cause
Cam Link not detected by Jetson — usually a loose USB connection at the hub or unpowered hub dropping the device.
Fix
Check USB connection at hub. Try a different USB port on the hub. Run lsusb on Jetson to confirm the Cam Link appears in the device list. Restart the agent: sudo systemctl restart rep1-agent
Cause
Poor lighting at the ID station, athlete not centered or too far from camera, or insufficient enrollment photos.
Fix
Add a battery-powered LED ring or panel facing the athlete's face. Ensure athlete stands 18–24 inches from the iPhone. Re-enroll using 5 selfie angles (not just 3). Check that the iPhone lens is clean and the room is not backlit.
Cause
GoPro battery below 15% (automatically reduces framerate to conserve power), or Media Mod HDMI port not fully seated.
Fix
Swap the GoPro battery. Re-seat the Media Mod by firmly pressing it until it clicks. If choppy behavior persists, reduce recording to 120fps — run sudo systemctl restart rep1-agent after changing settings.
Cause
Too many simultaneous AI inference models running, or the Jetson is in direct sunlight without ventilation.
Fix
Enable the fan and max clocks: sudo jetson_clocks --fan. Reduce load by disabling face recognition for non-enrolled athletes (ArUco-only mode). Shade the Jetson from direct sun with a small canopy or cloth. Target operating temp is below 70°C.
Cause
Marker obscured by jersey wrinkles, viewed at too sharp an angle, printed too small, or adhesion has loosened.
Fix
Use markers at least 4"×4". Re-attach with fresh Velcro and press flat. Run detection test: python3 /opt/rep1/tools/aruco_test.py. Ensure cameras are perpendicular and markers are clearly visible at 3–15 ft range. Avoid shiny laminate if detection rate is low.
Cause
WiFi signal too weak at the Jetson location, or the Azure Blob Storage endpoint is temporarily unreachable.
Fix
Move the router closer to the Jetson, or run an ethernet cable from the Jetson directly to a mobile hotspot. Test Azure connectivity: curl https://strep1prototype.blob.core.windows.net/. Use "Force Sync All" in the Upload Queue if clips are stalling individually.
Cause
Cosmos DB write lag (usually resolves in 30–60 seconds), or the session auth token has expired mid-event.
Fix
Hard refresh the athlete profile page (Cmd+Shift+R / Ctrl+Shift+R). Re-approve the clip from the queue. If profiles continue to not update, check Admin Console → System Health → Azure to see if the Cosmos DB connection is healthy. Re-login if the auth token indicator shows expired.
Cause
Incorrect camera zone assignments, uncalibrated distance markers, or cameras not precisely perpendicular to the athlete path.
Fix
Verify zone assignments in Admin Console → Camera Zones. Re-run AI calibration using a known 4.5s test run. Confirm each camera is exactly perpendicular (use a phone compass if needed). Acceptable AI vs. manual variance is ±0.05 seconds — anything larger indicates a calibration or zone issue.

8

Quick Reference Card

Print this card, laminate it, and keep it at the event table for fast operator reference.

Rep 1 Exposure — Field Reference Card Print & Laminate · v1.0
WiFi NetworkREP1-FIELD
WiFi Passwordrep1field2026
Admin Console URLrep1-admin.local
Jetson IP Address192.168.1.100
Start agentsudo systemctl start rep1-agent
Check agent statussudo systemctl status rep1-agent
Restart agentsudo systemctl restart rep1-agent
Enable fan (overheating)sudo jetson_clocks --fan
Power off Jetsonsudo shutdown -h now
Upload force-syncAdmin Console → Upload Queue → Force Sync All
GoPro settings1080p · 240fps · Protune ON · Stabilization OFF
Camera zonesStart, 10yd, 20yd, 40yd, Overview, Check-in
Min ID confidence75% (flags below for manual review)
Support emailsupport@rep1.io
Sources & References
1.GoPro HERO 13 Black specifications and Media Mod HDMI requirement — gopro.com/en/us/cameras/hero13-black
2.NVIDIA Jetson Orin Nano Super pricing and specifications — blogs.nvidia.com/blog/jetson-generative-ai-supercomputer/
3.Elgato Cam Link 4K product page — elgato.com/us/en/p/cam-link-4k
5.MICRO-TRACK multi-camera re-identification research — arxiv.org/html/2409.03879v1
6.iPhone 240fps timing validation for sprint analysis — pmc.ncbi.nlm.nih.gov/articles/PMC9003053/