STMicroelectronics’ Multi-Zone Time-of-Flight Sensor Enhances Spatial Awareness in Embedded Systems

Consumer electronics, smart building equipment, industrial machinery, and mobile robots all need to understand not just whether something is nearby, but where it sits within a scene and how far away it is. For example, a robot vacuum that can only detect an obstacle directly ahead will still collide with a chair leg approaching from the side. A smart display that wakes when a hand waves in front of it but cannot tell whether that hand swiped left or right also makes for a frustrating user experience. A building automation sensor that reports occupancy as a binary output cannot distinguish between one person seated at a desk and three people sitting around a conference table.

Standard proximity sensors fall short of these demands in multiple ways. For example, infrared reflectance sensors return a signal proportional to reflected IR intensity, but their readings vary with target color, surface finish, and ambient light conditions, making them unreliable without extensive per-application tuning. Single-zone Time-of-Flight sensors can measure an absolute distance that is independent of target reflectance, but they collapse the entire field of view into one distance value and therefore cannot localize objects spatially. Ultrasonic rangefinders work across a wider set of surface materials but have relatively slow update rates, wide beam angles that make it difficult to pinpoint small objects, and blind zones at close range.

STMicroelectronics’ VL53L8CX sensor combines Time-of-Flight distance measurements with multi-zone spatial resolution in a single miniature module, providing a low-resolution depth map of the scene in front of the sensor.

Figure 1: STMicroelectronics VL53L8CX sensor. (Image source: STMicroelectronics)

Multi-zone Time-of-Flight changes the design equation

Time-of-Flight sensing works by emitting a short pulse of infrared light from a vertical-cavity surface-emitting laser (VCSEL) and measuring how long it takes for that light to return after reflecting off objects. Since the measurement depends on photon travel time rather than reflected signal intensity, the resulting distance reading is absolute and remains consistent regardless of a target’s color or surface reflectance, unlike reflectance-based sensors that must be calibrated for each target surface.

The VL53L8CX expands this principle from a single distance point to either a grid of 64 independent measurement zones in an 8 x 8 matrix, or 16 zones in a 4 x 4 matrix, across a 45° x 45° square field of view (65° diagonal). Each zone returns its own distance reading, and the sensor's patented histogram algorithms enable it to detect and track multiple objects at different distances within the same zone. If two objects sit at different distances along the same line of sight (a wire-mesh fence in front of a wall, for example), the histogram will show two distinct peaks, and the sensor reports both distances rather than averaging into one ambiguous value.

The VL53L8CX’s histogram processing offers excellent cover-glass crosstalk immunity beyond 60 cm, which simplifies housing design for consumer products where the sensor is behind a protective window. The result is a low-resolution depth map, updated at rates up to 15 Hz in 8 x 8 mode or up to 60 Hz in the faster 4 x 4 mode, that captures enough spatial information for the host microcontroller to make zone-based decisions about obstacle locations, occupant patterns, gesture direction, or container fill levels. The 940 nm VCSEL emission is invisible to the human eye and carries a Class 1 laser safety certification, allowing the sensor to be utilized in consumer products without safety enclosures or warning labels.

Applications configurations for robotics, building automation, and gesture control

The VL53L8CX communicates over both I²C (up to 1 MHz) and SPI (up to 3 MHz), with the SPI option useful for applications that require fast host data transfers and reduced initialization time. This is a huge advantage, since the sensor loads approximately 84 KB of firmware at startup, a transfer that SPI completes considerably faster than I²C. Moreover, since the sensor handles all histogram processing and distance computation internally, a host MCU receives pre-processed zone distance data rather than raw photon counts, which helps to keep the computational load low enough that even a modest Cortex-M0+ or Cortex-M4 class device can manage the sensor alongside other system tasks.

In a mobile robot or autonomous vacuum cleaner, mounting the VL53L8CX on the front or side of the chassis can provide an 8 x 8 depth grid covering the forward path. Zone-based obstacle detection can also distinguish between a wall (filling multiple zones at a consistent distance), a narrow chair leg (appearing in one or two zones against a more distant background), and where a floor drops off, like a staircase (where the lower zones suddenly return to longer distances). Additionally, cliff detection, small-object avoidance, wall tracking, and floor-type recognition can all be implemented with zone-comparison logic on the host MCU, with a frame rate up to 15 Hz in 8 x 8 mode or up to 60 Hz when the application can operate in 4 x 4 mode. This feeds the control loop updated scene data at a rate appropriate to the chosen configuration to adjust heading and speed in real time.

Figure 2: An autonomous vacuum cleaner. (Image source: Adobe Stock)

Building automation and retail environments are also potential use cases that require a different configuration. Here, mounting the VL53L8CX on a ceiling or above a doorway and pointing it downward can turn the 8 x 8 grid into a presence and occupancy map of the floor below. Since the sensor measures absolute distance and is unaffected by clothing color or ambient lighting variation, it can reliably count the number of people in a zone, track movement direction for entrance and exit counting, and distinguish between occupied and empty desks for controlled ventilation and lighting. Moreover, its low-power mode and programmable distance threshold also allows the sensor to remain in a sleep state and wake the host only when someone enters a defined range.

As part of STMicroelectronics’ STGesture platform (which includes the STSW-IMG035 turnkey gesture-recognition software and the Gesture EVK development tool), the VL53L8CX can also provide the spatial precision needed for gesture-based interaction with smart displays, lighting controllers, and other appliances. Pairing the multi-zone distance data with AI models from the STM32ai-modelzoo on GitHub can extend the interaction vocabulary beyond simple directional swipes to include “FlatHand,” “Fist,” and “Like” gestures drawn from eight hand-posture classes in ST’s pre-trained VL53L8CX hand-posture dataset.

What designers need to know before integrating the VL53L8CX

Sensor placement and field-of-view alignment are some of the most crucial considerations in any VL53L8CX integration. The 45° x 45° square projection defines the detection area at a given distance, so designers can model the coverage footprint for their mounting height and angle early in layout to confirm that the zones of interest align with the regions where detection is required. The VL53L8CX’s cover-glass crosstalk immunity, consistent beyond 60 cm by the histogram algorithms, improves optical housing designs by allowing the sensor to operate behind a shared cover glass without using an isolated optical window, though the glass opening must meet or exceed the receiver exclusion zone dimensions specified in the datasheet to avoid clipping the outer zones.

On the software side, STMicroelectronics provides a free driver, API, and evaluation software ecosystem that drastically reduces the integration time. The P-NUCLEO-53L8A1 evaluation kit pairs the sensor with an STM32 Nucleo board for rapid prototyping, and the reference software can be migrated to custom hardware with minimal modifications since the driver abstracts the I²C/SPI transport layer. The sensor’s built-in motion indicator reports per-zone motion intensity values, which allows the host to detect that a target has moved and, by tracking which zones register activity across sequential frames, determine the direction of that movement.

Coming soon: the VL53L9CX

Teams building around the VL53L8CX today will also want to keep an eye on what comes next. STMicroelectronics has announced the VL53L9CX, a next-generation 3D lidar camera module expected to become available in early-to-mid 2026. Not just a modest iteration, the VL53L9CX will provide scaling to up to 2,268 zones (54 x 42), a 9-meter ranging capability, dual VCSELs, an integrated postprocessing SoC, and combined depth and 2D IR image output in a larger module. Although the VL53L8CX and VL53L9CX are not software or hardware-compatible, the ToF system design experience and familiarity with STMicroelectronics’ FlightSense ecosystem from VL53L8CX development will serve as a good foundation for evaluating the newer device.

Conclusion

Going from single-point distance measurement to multi-zone spatial awareness has been one of the most noteworthy improvements in embedded sensing in recent years. Compressing what amounts to a 64-pixel depth camera into a 6.4 x 3.0 mm reflowable module that draws modest current, runs at 60 Hz, and outputs data over a standard serial bus, allows the VL53L8CX to provide the spatial intelligence needed to build smarter robots and more responsive human-machine interfaces without the cost, complexity, or power overhead of camera-based or scanning lidar solutions.

For additional information on the VL53L8CX’s specifications and other resources, please visit the VL53L8CX Time-of-Flight Sensor product page.