Scentience designs and manufactures neuromorphic sensors, fusion software, and AI for olfaction applications in robotics. We became frustrated with the current state of olfaction sensors as none of them are optimized for real-time robotics. So we built our own, optimizing the entire hardware and software engineering stack in the process. Our custom silicon, multi-patent-pending OPUs are the heart of each device. All devices are currently for research purposes only and not for sale, with a few close industry and university partners working with us to build large and standardized olfactory datasets for robotics. However, we are always happy to discuss our awesome tech with potential partners who share our vision of bringing the sense of smell to machines. If this sounds like you, please drop your email below. Industrial partnerships by invitation only. Scentience devices are designed to meet the FCC Part 15 requirements - full FCC testing in progress with MIL-STD-810H testing to follow.
Reconnaisscent is designed for rapid, sub-second chemical sensing in robotics applications at the edge. Slightly bigger than a bottle cap, Reconnaisscent contains a single olfactory processing unit (OPU) with the ability to detect 16 different (and programmable) chemical compounds along with temperature, barometric pressure, and humidity. Also on board the device is an IMU for better scent-based navigation through olfactory inertial odometry.
The Scentience APIs enable developers to communicate with the device over BLE, Wi-Fi, I2C, SPI, and USB to stream live chemical data to their desired platform.


The small form-factor enables the device to be used on the lightest of UAVs, humanoid robots, and other edge devices to perform scent-based navigation. Navigation purely by scent is now real and repeatable, and we were the first to prove it.

Swarm intelligence enables many devices to be networked together to triangulate the source of chemical compounds within a building or robot training environment. Each device comes pre-installed with the necessary networking software.
Scentinel is designed for more qualitative analysis of air samples for comprehensive chemical composition. With higher chemical resolution (part-per-trillion) than its Reconnaisscent counterpart, Scentinel houses 8 OPUs for the detection of hundreds of chemical compounds. Onboard multimodal AI works with the Scentience App to perform long-context chemical analysis.
Machine olfaction lags behind other sensory modalities partially due to the lack of large training datasets. To solve this, Scentinel was designed to operate in laboratory settings for the acquisition of mass amounts of training data for multimodal AI models.


Scentinel was designed to acquire mass amounts of multimodal training data for AI foundation models at research institutions. The device operates with the same seamless APIs and AI models as Reconnaisscent.

Field tests with Scentinel have been performed within botanical and aerospace environments to infer results in long-context chemical patterns that can inform safety, air quality, and business decisions. Data is accessed via Scentience dashboard or mobile app..
To encourage progress in machine olfaction for robotics, Scentience offers an olfactory development board that is equipped with optical, acoustic, electrochemical, and metal oxide sensors. These boards are equipped with the base code to communicate with the Scentience App and APIs, and are intended for research purposes only.
Boards are expected to be ready for distribution to industry partners in January 2026.
If you would like the opportunity to be an early adopter of our development boards, please join the waitlist below.

Copyright © 2024-2025 Scentience Robotics - All Rights Reserved.