Part of my training for doing "engine room checks" on a boat involved checking for any unusual smells, e.g. fuel leak, burning oil (from generator/engine), burning coolant (from generator/engine), or burning rubber (from sea chest raw water impeller). All of the components in there are equipped with sensors[1] that measure levels, temperature, etc. Perhaps there is room for a new olfactory sensor there? Aside from avoiding catostrophic issues like fire and engine or generator failure, it's also important to not pump out[2] any water from the compartment into the ocean if it's contaminated with oil, fuel, or coolant (the laws about this are super strict).
[1] There are digital sensors that are readable directly from the pilothouse by the captain which are rigged to automated alarms, as well as manual sensors (e.g. a pressure dial) that are readable from the engine room itself, for redundancy. So I don't think an olfactory sensor would replace the unusual smell check, but it could maybe augment it.
[2] The "bilge pump" is used to pump out water from the bilge (bottom floor cavity of engine room). To be honest on my vessel the policy is to never turn on the bilge pumps in the engine room at all because the risk of dumping contaminants is too high. But I still thought to mention this just in case there's an idea there.
I have a friend with Chrons, IBS, and a handful of other gut issues. He wants me to build something like this to help self-diagnose acute issues as they arise. Yes, a fart classifier.
I want to use a smell classifier to identify ripeness levels in agriculture.
I haven’t tested to see if this is even feasible, but I’d like to also use a tool like this for pest scouting in agriculture. If the sensors are sensitive enough to detect small amounts of fungi, arthropod activity, or hormonal shifts, this could be useful for early detection in integrated pest management systems.
We conducted research with local universities, and the digital nose was able to detect the presence of pests in oat flakes and beans (two different species).
Unfortunately, medical applications require enormous time and effort to meet strict verification and regulatory requirements. While this is an important long-term direction, we are currently focusing on lower-hanging opportunities such as food manufacturing and processing, where there is strong potential for cost savings and loss prevention.
Digital smellers are scalable and more repeatable than human noses. At the current stage our electronic nose operate either through classification of previously trained odor classes or through anomaly detection.
What is still missing is a possibility to run a more sophisticated conversation with the model when something smells "suspicious".
The problem is not whether we can digitize the sense of smell, but that no industrial process currently relies on it by default. The real challenge is identifying the first scalable use case that proves measurable business value (sniphi team member here).
The nearest current use of detection of particles in the air that I can think of is smoke and carbon-monoxide detectors for safety. Could adoption on these smart versions like Nest or Ring by adding your sniphi detector provide other types of early warning systems for safety, air quality or sensing?
Some thoughts are musty odors from mold/mildew, rotten egg smells indicating gas leaks, and fishy/burning plastic odors from electrical issues.
That is actually an interesting direction. Since smoke detectors already exist, the next level could be distinguishing smoke from a cigarette — or even something harmless like burning scrambled eggs — from more dangerous sources such as burning carpet or electrical wire insulation. We will definitely think about it.
A mold detector is also an interesting idea. Our ‘digital nose’ can measure humidity and temperature as well, and these factors are often strongly correlated with mold growth. Combining odor detection with environmental data could therefore be very useful for early mold detection.
That’s true. We even started a PoC with a skincare products factory. The challenge, however, was that the frequent rotation of the product portfolio — and the large number of SKUs — made it difficult to justify the training effort.
On the limits of detection - with Sniphi we follow a different approach than traditional selective sensors. The system is based primarily on non-selective chemical sensors operating at controlled temperature profiles. Each measurement cycle (6 seconds) generates around 60 measurement points per sensor, creating multidimensional signatures of gas mixtures that are then analyzed using classification models.
Interesting direction. My guess is the first strong wedge is narrow pass/fail decisions where people already use smell informally and misses are expensive: fermentation batches, packaging seal leaks, or early spoilage or mold detection in storage. If you can show earlier-than-human detection plus low recalibration burden across facilities and seasons, the ROI story becomes much easier to sell than a broad platform story. How close are you on handling humidity and temperature variation plus sensor drift without site-specific retraining?
That’s a very good point, and we actually see fermentation batches as one of the most promising early use cases. In many facilities, smell is already used informally as a pass/fail indicator, but it is subjective and difficult to scale.
We measure both humidity and temperature and use them as additional inputs for the ML models. Regarding sensor drift, it is still difficult to fully assess its impact on the business case. At this stage, our main focus is on the accuracy of the classification models rather than very long-term operation — that would be the next step.
For now, the practical approaches we consider are either on-the-fly calibration through a feedback loop based on the actual process output, or simply replacing the sensor when necessary, as the manufacturing cost is relatively low.”
From the pictures, it looks like it's using sensirion VOC sensor. There are plenty of "experimental" VOC detectors in the market, including BME688/690 with their AI SDK, so far I haven't seen a single reliable industry-grade application, only demos that work sterile conditions and fail in the harsh real-world conditions.
[1] There are digital sensors that are readable directly from the pilothouse by the captain which are rigged to automated alarms, as well as manual sensors (e.g. a pressure dial) that are readable from the engine room itself, for redundancy. So I don't think an olfactory sensor would replace the unusual smell check, but it could maybe augment it.
[2] The "bilge pump" is used to pump out water from the bilge (bottom floor cavity of engine room). To be honest on my vessel the policy is to never turn on the bilge pumps in the engine room at all because the risk of dumping contaminants is too high. But I still thought to mention this just in case there's an idea there.
I have a friend with Chrons, IBS, and a handful of other gut issues. He wants me to build something like this to help self-diagnose acute issues as they arise. Yes, a fart classifier.
I want to use a smell classifier to identify ripeness levels in agriculture.
I haven’t tested to see if this is even feasible, but I’d like to also use a tool like this for pest scouting in agriculture. If the sensors are sensitive enough to detect small amounts of fungi, arthropod activity, or hormonal shifts, this could be useful for early detection in integrated pest management systems.
When we published the white paper ( https://sniphi.com/wp-content/uploads/2025/10/Sniphi_digital... ), we expected a queue of agricultural companies interested in the technology. However, pests apparently aren’t “sexy” enough to capture attention.
We observed the same reaction with bananas — fresh vs. overripe, like in the video. Technically interesting, but no one saw clear business potential.
So now we are looking for use cases that are more obvious and compelling from a business perspective. Any ideas?
How good are digital smellers compared with super human smellers?
Digital smellers are scalable and more repeatable than human noses. At the current stage our electronic nose operate either through classification of previously trained odor classes or through anomaly detection. What is still missing is a possibility to run a more sophisticated conversation with the model when something smells "suspicious".
Some thoughts are musty odors from mold/mildew, rotten egg smells indicating gas leaks, and fishy/burning plastic odors from electrical issues.
A mold detector is also an interesting idea. Our ‘digital nose’ can measure humidity and temperature as well, and these factors are often strongly correlated with mold growth. Combining odor detection with environmental data could therefore be very useful for early mold detection.
What is the limit of detection on the sensors? Can they reliably pick up compounds in the parts per billion range? Parts per trillion?
On the limits of detection - with Sniphi we follow a different approach than traditional selective sensors. The system is based primarily on non-selective chemical sensors operating at controlled temperature profiles. Each measurement cycle (6 seconds) generates around 60 measurement points per sensor, creating multidimensional signatures of gas mixtures that are then analyzed using classification models.
We measure both humidity and temperature and use them as additional inputs for the ML models. Regarding sensor drift, it is still difficult to fully assess its impact on the business case. At this stage, our main focus is on the accuracy of the classification models rather than very long-term operation — that would be the next step.
For now, the practical approaches we consider are either on-the-fly calibration through a feedback loop based on the actual process output, or simply replacing the sensor when necessary, as the manufacturing cost is relatively low.”