indie Semiconductor has released the iND881, a system-on-chip (SoC) for edge AI in automotive and robotic applications. The chip integrates AI compute with a multi-camera image signal processor (ISP) to support smart camera systems.

The iND881 architecture includes a Neural Processing Unit (NPU) for AI workloads, a Digital Signal Processor (DSP) for classical signal processing, and a quad-core ARM Cortex-A53 CPU. It features a low-latency H.264 encoder for multi-channel video streaming and an ISP capable of processing multiple camera inputs with sub-1 ms latency.

The SoC is designed for automotive safety features, including driver and occupant monitoring systems (DMS/OMS) and electronic mirrors with blind-spot detection. It also supports sensor modalities such as infrared, thermal, time-of-flight, Radar, and LiDAR, making it applicable for autonomous mobile robots and humanoid systems.

The device is ASIL-B compliant and includes a hardware security module (HSM) for cybersecurity. It can be paired with perception software from emotion3D to provide a pre-integrated hardware-software package for vehicle manufacturers and Tier 1 suppliers.

“With the launch of iND881, we are realising opportunities in edge AI markets such as automotive and humanoid robotics,” said Fred Jarrar, senior vice president at indie. He stated that the platform offers customers flexibility in their design approach for automated sensing.

The iND881 is currently sampling to customers. The company plans to demonstrate the technology at the AutoSens and InCabin USA 2026 exhibitions.