Syntactica Ltd
Robotics research • Hardware systems design
Dual-spectrum robotic vision

eCam32
Dual-Spectrum Robotic Vision

A mobile visible + NIR (Near-Infrared) 5MP imaging platform featuring Wifi, Bluetooth & bi-directional audio. Based on the dual-core Espressif ESP32-S3 32-bit RISC processor.

Hardware trigger + timestamping
Deterministic sync
SDK + reference drivers
Visible-spectrum demo frame
End-to-end latency: < 20 ms (target)

Overview

This page is laid out to “sell” to technical buyers without marketing fluff: quick positioning + credible engineering claims, then deeper specs, block diagrams, and integration documents.

What it is

A robotics-focused imaging PCB integrating two camera paths: visible and NIR (Near-Infrared). Designed to support deterministic capture via hardware trigger, coherent timestamping, and a clean host interface for low-latency perception pipelines.

Active IR illumination Low-light robustness Triggered capture Frame-accurate sync

Specifications

Use a compact table: it’s skimmable and forces you to state measurable numbers. Replace “TBD” with verified values.

Key specifications
Spectral bandsVisible (RGB) + NIR (Near-Infrared), e.g. 850–940 nm (TBD)
Resolution / frame ratee.g. 1920×1080 @ 60 fps per channel (TBD)
Shutter typeGlobal shutter preferred for robotics (TBD)
SynchronizationHardware trigger input, shared clocking, frame counters (TBD)
TimestampingHost-visible timestamps; PTP (Precision Time Protocol) option if Ethernet (TBD)
Host interfaceUSB 3.x / MIPI CSI-2 / GigE (Gigabit Ethernet) (choose)
Power inpute.g. 9–24 V DC (Direct Current) with onboard regulation (TBD)
Typical powere.g. 3.5 W @ nominal capture (TBD)
Operating temperaturee.g. −20 to +70 °C (board) (TBD)
EMI/EMCDesigned with ground/return integrity and shielding options (TBD)

Acronyms: fps = frames per second. PTP = Precision Time Protocol. EMI/EMC = Electromagnetic Interference / Compatibility.

Electrical interface notes
  • Trigger: 3.3 V logic input, configurable edge, debounce/filtering optional (TBD).
  • GPIO: status pins for frame-valid, sync, error state (TBD).
  • Power: reverse polarity and inrush handling recommended for robotics harnesses.
  • Grounding: star-ground strategy into chassis + controlled return paths for high-speed lanes.
Optical considerations
  • IR illumination: 850 nm offers sensitivity; 940 nm reduces red-glow visibility (TBD).
  • Filters: band-pass filters reduce cross-talk between visible and NIR paths.
  • Calibration: intrinsic/extrinsic calibration for dual-path alignment (factory or user workflow).
Latency & determinism
  • Capture-to-host: quantify camera exposure + readout + transport + host copy.
  • Timestamp point: define whether timestamp is at trigger edge, exposure start, or frame-valid.
  • Jitter budget: publish worst-case timing jitter under load (engineers care).

Block schematic diagram

For the website, a clean block diagram is usually better than dumping full schematics inline. Link the full schematics PDF above.

Imaging Sensor 5MP (2592 × 1944 pixels) Selectable IR Filter USB / Programming USB D+/D- to ESP32-S3 USB EN + IO0 bootstrap / reset Optional UART bridge Audio Input Mic Input + Preamp ADC path ESP32-S3 Wifi + Bluetooth LE module Dual Core RISC USB OTG OTA updating Vector instructions 16MB Flash QSPI (internal) Boot + App 8MB PSRAM QSPI (internal) Frame buffer / ML Audio Output I²S to integrated Class D Amp Speaker connection Stepper Drive Darlington output Supports 2 x 5-wire steppers Illumination 2 x White LEDS 2 x NIR LEDS Ambient light sensor Daughter board Storage Integrated microSD slot SPI Interface Power Selectable supply 9-12v / 5v / USB Vbus Reverse protected 2.4 GHz Optional U.FL connection Block-level only. Full schematics: see the PDF above.
OTG = On The Go; OTA = Over The Air; NIR = Near Infra-Red
Integration checklist (example)
  • Trigger semantics: specify what the trigger edge means (exposure start vs frame latch).
  • Timebase: define how timestamps are generated and how to align with robot clock.
  • IR illumination: specify supported wavelength(s), strobe support, and safety constraints.
  • Calibration: publish a repeatable workflow for visible↔NIR alignment.
  • Thermals: provide heat paths, derating, and enclosure guidance.

Use-cases

Keep use-cases concrete. Tie each to a measurable benefit (lighting robustness, timing determinism, failure handling).

Adverse-lighting perception
Visible degrades under glare, haze, or low light; NIR with active illumination improves feature stability.
Low light Glare Active IR
Deterministic sensor fusion
Hardware-triggered capture reduces jitter in multi-sensor stacks (IMU (Inertial Measurement Unit) + encoders + cameras).
Trigger Timestamp Fusion
Inspection and verification
Dual-band imaging helps detect materials/inks/markings not reliably visible in RGB alone.
Inspection QA Registration

Evaluation kit & support (example)

This is where you state what the customer gets and what you support. Keep it crisp and testable.

What ships (example)
  • Vision PCB + lens/sensor modules (as configured)
  • Harness + trigger cable
  • Reference host capture utility + SDK examples
  • Calibration target + procedure

Keep claims consistent with what you can actually ship/support.

Contact / purchase (dummy)

Replace with real addresses/forms. Avoid putting personal email addresses on public pages if possible.

© Syntactica Ltd
Dual-Spectrum Vision PCB — example layout and content scaffolding.