Sintrones IBOX-601-G2-ON4 NVIDIA Jetson Orin Nano 4GB In-Vehicle Edge AI Computer with 4x GMSL-2 Cameras

Proving a vehicle vision idea across a fleet is mostly about cost discipline: enough compute to run the model, the same cameras and wiring as the production design, and nothing spent on headroom the pilot will not use. The IBOX-601-G2-ON4 is the entry build of the Sintrones IBOX-601-G2 series, an in-vehicle edge AI computer running an NVIDIA Jetson Orin Nano 4GB module behind the four GMSL-2 camera inputs shared across the family.

The Orin Nano 4GB suits single-model workloads: lane or object detection on a forward camera, driver monitoring, automatic recording triggers, or capture-and-forward duty where most inference happens elsewhere. Because the I/O and enclosure are identical across the series, a pilot built on the IBOX-601-G2-ON4 can move to the IBOX-601-G2-ON8 or an Orin NX build without touching brackets, looms or camera choices.

Camera capture uses the same four GMSL-2 inputs on FAKRA-Z connectors found across the range. Networking offers 2.5GbE and GbE on M12 X-coded connectors, with CAN FD and RS-232/422/485 on DB-9 for vehicle integration, two USB 3.2 ports and an HDMI output. An M.2 2280 NVMe SSD handles the system image and local storage, and M.2 Key B with a nano SIM slot, M.2 Key E and mPCIe slots add optional 4G or 5G, Wi-Fi and GPS.

Vehicle power is handled by a 9-60 V DC M12 K-coded input with smart power and ignition management, and the in-vehicle edge AI computer runs fanless from -25 °C to 70 °C with MIL-STD-810G shock and vibration testing. E-Mark and EN 50155 compliance let the IBOX-601-G2-ON4 serve road and rail programmes, and Sintrones rates storage from -40 °C to 80 °C for parked vehicles in extreme climates.