IN Brief:
- The 1.1MP VD56GA targets driver and occupant monitoring across higher-volume vehicle platforms.
- ST reports 35% higher modulation transfer function and nearly 60% higher quantum efficiency than its previous generation.
- Samples are available now, with front-end manufacturing carried out at Crolles in France.
STMicroelectronics has introduced the ST SafeSense VD56GA, a 1.1MP automotive image sensor designed to reduce the cost and integration burden of infrared driver and occupant monitoring. The device combines a compact 3.7mm × 3.2mm chip-scale package with higher near-infrared sensitivity and embedded image-processing functions for high-volume vehicle platforms.
The sensor uses a backside-illuminated pixel architecture that ST calls DeepNIR. Compared with the previous generation, the company reports a 35% improvement in modulation transfer function and almost 60% higher quantum efficiency, increasing infrared response without moving to a larger sensor format.
ST is addressing camera-module cost alongside sensor performance. Higher infrared sensitivity can reduce the illumination requirement, while improved optical performance can permit lower-cost lenses and less demanding filtering. The VD56GA also integrates image-processing functions that reduce the need for additional external processing inside the camera module.
Those changes are relevant to camera architectures where space, power, component count, and thermal load are tightly constrained. A monitoring camera still has to interface with a domain or central processor, but carrying out more of the basic image preparation on the sensor can simplify the electronics between the pixel array and the vehicle’s higher-level perception or monitoring software.
Driver monitoring has expanded beyond premium vehicles as manufacturers introduce systems intended to track attention, eye position, head movement, and other indicators. Occupant monitoring extends the same imaging hardware towards functions such as seat occupancy, passenger position, and cabin observation, increasing the incentive to use a common sensor architecture across several vehicle classes.
Infrared performance is particularly important because cabin cameras must continue operating through changing daylight conditions and at night. Improving quantum efficiency gives module designers more freedom over infrared-emitter power and placement, although the complete system still has to balance illumination, optics, sensor noise, thermal behaviour, and reflections from glasses, trim, and display surfaces.
The compact CSP also creates more placement options, including installations where the camera has to sit behind or beneath another cabin surface. Such positions impose their own optical penalties, so sensor sensitivity and modulation transfer performance become more useful when they allow the wider optical stack to tolerate lower transmission or more constrained geometry.
The VD56GA extends an automotive imaging portfolio that ST says has already passed 10 million SafeSense sensor shipments. The company is also developing imaging technology that places more perception capability towards the edge, including sensor and depth-processing work intended to reduce reliance on central processing for every imaging task.
Manufacturing is split between ST’s front-end operation at Crolles in France and downstream assembly in Asia. Automotive programmes place unusual weight on production continuity because design cycles and qualification periods are long, while changing a camera sensor can affect optics, software, calibration, functional-safety work, and regulatory validation.
That manufacturing route consequently forms part of the product proposition. ST controls the sensor’s front-end fabrication inside its own semiconductor network, giving vehicle manufacturers and Tier 1 suppliers a defined production source for programmes that may remain active for many years.
Market volume is expected to rise as cabin sensing appears across more vehicle segments. ST cites Yole Group forecasts indicating driver and occupant imaging volumes could exceed 70 million units by 2031. At that scale, the engineering competition moves beyond maximum image quality towards the combination of optical performance, package size, illumination cost, processing partition, and manufacturing stability.
Samples of the VD56GA are available now, with pricing and sample requests handled through ST sales offices. Its specifications reflect the direction of volume automotive sensing: enough infrared performance to support demanding cabin positions, while reducing the supporting optics and electronics that determine whether the camera can be economically fitted across an entire vehicle platform.
