What is Sniphi?
Sniphi, a technology start-up founded by Antdata, is a leader in digital scent recognition.
The next milestone in the AI and robotics revolution is scent. Until now, machines have learned to see, hear, and speak—but they haven’t been able to smell. We created Sniphi, the digital nose, to change that.
How Sniphi works?
The Digital Nose’s sensors capture real-time data on the unique patterns of gases and VOCs (Volatile Organic Compounds), which are processed by the AI-powered digital brain. The system identifies these patterns, enabling real-time analysis and providing accurate insights about the detected scent or gas.
What is a DIGITAL NOSE?
The Digital Nose is a modular platform adaptable to various industries. It consists of:
ADVANCED IOT SENSORS
AI-POWERED DIGITAL BRAIN
FLEXIBLE DATA ARCHITECTURE
IMPLEMENTATION PROCEDURE
Our sensors provide real-time data on the unique patterns of gases and VOCs (Volatile Organic Compounds), which serve as inputs for the AI-powered digital brain.
The system uses machine learning (ML) algorithms to recognize these patterns, enabling real-time analysis and delivering precise information about the detected scent or gas.
This is made possible through IoT-class terminals and cloud-based Microsoft services like IoT Hub, Databricks, Power Apps and Power BI.
The implementation of the Digital Nose platform follows a standardized procedure and can be easily integrated with the customer’s existing Microsoft environment or deployed as intelligent sensors on an edge device (AI on the Edge).
How it works?
During the model training phase, data from the digital nose sensors is seamlessly transmitted to the Microsoft Azure cloud (our “Digital Brain”), where it can be:
- Analyzed in real-time with Power BI.
- Stored in an Azure SQL Database.
- Labeled for AI model training using Power Apps.
- Processed by Machine Learning algorithms using Azure Databricks.
Once the model is trained, it analyzes the "digital fingerprint" of gas or scent data and compares its profile with a database of known fingerprints. This enables instant identification and provides detailed insights about the measurement — which can also trigger automated actions.
The pre-trained model can be deployed either directly on the sensor (AI on the Edge) or hosted in the cloud and accessed via API requests (AI in the Cloud).
Applications
Since each substance emits a unique pattern of VOCs and gases, Sniphi's Digital Nose can create a "digital fingerprint" of each substance, analyze it with AI, and display the results in Power BI.
Thus, our Digital Nose platform can address challenges across multiple industries:
- Food and Beverage: Monitors product freshness and spoilage, ensuring quality control.
- Safety and Security: Detects hazardous gases and chemicals to improve workplace safety.
- Fragrance and Cosmetics: Profiles scents and supports new product development.
- Medicine: Helps to identify biomarkers and monitor the stability of high-risk processes.
- Environmental Monitoring: Identifies pollution and harmful substances.
And many other industries.
Implementation
The Digital Nose integrates seamlessly with Microsoft Services, making it easy to implement and integrate with a customer's existing Microsoft tenant.
The standardized implementation process includes:
- Configuration of the required Microsoft services.
- Installation of plug-and-play PoE sensors and the IoT terminal.
- Deployment of the trained model — either on the sensor (AI on the Edge) or in the cloud (AI in the Cloud)
- Power Apps-based AI model training.
- Development of tailored Power BI reports.
This standardized approach allows the digital sensors to be installed almost anywhere, from manufacturing lines and warehouses to laboratories and offices.
Moreover, there is no need for a large initial investment to test our Digital Nose. You can begin with a limited-scope Proof of Concept (PoC) and, after successful implementation, expand the solution.