A distributed network of sensors to warn people about allergens
Image credit: Image credit: Wolfgang Kumm/dpa
Subash ChapagainApr 14, 2022
Please leave the feedback on this idea
Originality
Is it original or innovative?
Feasibility
Is it feasible?
Necessity
Is it targeting an unsolved problem?
Conciseness
Is it concisely described?
Bounty for the best solution
Provide a bounty for the best solution
Bounties attract serious brainpower to the challenge.
A distributed network of various types of biological and environmental sensors that proactively warns prone individuals about the different allergens circulating in the environment in real-time.
Why?
In the US only, allergies are the 6th leading cause of chronic illnesses, with an estimated economic cost of $18 billion annually. More than 50 million Americans suffer from allergies of different kinds .
A system that could detect allergen levels in real-time and sends warnings to people on their phones will help them guard against the allergen proactively, massively helping in well-being and health burden reduction.
How does the system work?
Basically, the system will be something like distributed weather stations. The only difference is that there will be more advanced sensors that can detect both biological and environmental allergens, and they will be more frequently distributed depending on the population/environmental conditions.
Types of sensors used:
a) Biological sensors
These sensors will be used to detect levels of specific compounds/molecules/combination of compounds that are responsible for allergic reactions. In fact, there are already some developments in this direction . These devices use simple biological markers sensing approach coupled with machine learning.
b) Abiotic sensors
These sensors are mostly similar to sensors used by smart weather stations. They collect data on dust, smoke and particulate levels; moisture; UV radiation and similar environmental features associated with allergic reactions.
Once the sensors collect the sensing data, they report the data to a LOCAL server/database in each specific region. To determine where to set up these local servers is a pretty straightforward task: we just need to find out areas with the highest number of allergic individuals, and areas with indicators of stochastic weather patterns (for eg, eczema allergy is triggered when sudden temperature fluctuations occur).
The servers will then send out trigger warnings to people based on their locality. With time, the sensors will be able not just for real-time warnings, but for predictions as well. For instance, for the past 3 years, flower blossoms in this ABC location caused these PQR insects to swarm into the area causing XYZ kind of allergy, so beware this spring in March-April!
How does this help?
Pretty simple: If I know I might have an eczema flare-up later today because the weather is going to be crazy dry, then I could moisturize myself properly before I leave my home. If I know there's a lot of pollen flying up in the air outside, I could wear a mask and take some antihistamine tablets with me. I just need to turn my notifications on!
From all the particles that are sampled from the air, could we somehow automatically identify which species they belong to? Then output a percentage of particles per species per day per area.
People could then record (on the matching app) on which days they felt the symptoms and check what was in the air that day. This eventually helps people learn what they should avoid.
An AI-powered prediction algorithm can take into consideration various factors such as temperatures, length of daylight, rainfall, wind, etc. to predict when such days will repeat.
Please leave the feedback on this idea
jnikola2 years ago
A company called BreezoMeter did exactly that! They give you information based on three plant types (weed, grass, and tree), but in some areas can go down to a plant-subtype level. However, they did it by satellite imaging and AI.
Doing it with an in-place monitoring device would be extremely heavy, as Subash Chapagain mentioned. I am keen on believing it would be done by protein recognition rather than gene sequencing, but yes, if you wanted to go to a detailed plant species level, then yes, it would be technologically and computationally exhausting. Also, putting these devices in the perfect location to provide precise information could be tricky. I bet on satellites and the power of AI!
Please leave the feedback on this idea
Subash Chapagain2 years ago
This could be possible but probably will require much-advanced biosensing. The identification at the species level usually would need much detailed analysis (will have to look at the gene sequence itself most probably) because normal molecule level biosensing uses broad biochemistry in general.
Alternatively, what can be done is to collect the samples physically from the sensing locations in a row, sequence the ones associated with the days when a higher incidence of allergenicity was reported, and then in the future train the algorithm to associate the gene sequence with that particular biochemical/biomolecular pattern. In this way, the system should be able to resolve the species-level difference after a few iterations.
Please leave the feedback on this idea
Car-mounted real-time pollen detecting device
jnikolaMay 02, 2022
Using a car-mounted pollen detecting device and a WIFI network to share real-time pollen data to users.
Why?
Increased precision due to many interconnected and moving detectors on the street level
How would it work?
The users would install a small detector design to fit inside the ventilation mask at the car front. The device would be powered by the car battery and would connect to a smartphone via Bluetooth. It would use the phone's location data to embed it with the pollen sample.
The sampling and analysis
The device would have two parts - a pollen collector and a camera. The pollen collector would be filter-based. While the car is driving, filter would collect particles. Every hour or a minute (depending on the efficiency of collection), the camera would take a detailed picture of the filter. Since filter would be coated with different dyes7antibodies that are specific to certain proteins found in differentkinds of pollen, the camera would detect those signals. Taking photo every once in a while would allow to eliminate the previous signals and connect the current picture to a certain location/area. These photos along with the location tags would be sent to the core facility via the app. AI would analyze these photos and give users of the app the real-time data on airborne pollen based on their location. The user would get a small compensation for being a "sampler"., while the users using the app would pay a small fee.
Add-on functions
The app would also have premium functions to help you plan a route with smallest concentration of pollen, detect the type of pollen that causes your symptoms, etc.
Additional information
Most of the pollen sensing and pollen maps are currently created by the integration of satellite imagery and forecast data (wind speed, precipitation, temperature, etc.). The companies leading the research in that field state that they can be precise in a 1 km range. However, to make the pollen prognosis more precise and personal for each user, we must go into more detailed pollen tracking and reporting. That's the reason why your idea is great!
Current methods of "local" pollen detection
SensioAir - mentioned by Subash Chapagain
Burkard spore trap, Hirst trap, Rotarod system (silicone grease-coated clear rod)- capturing airborne pollen on a glass slide, staining and counting on a microscope; measurements twice a day, every day; information about types of pollen
Automated Airborne-Particle Sensors - Acquire pollen, mold and dust data automatically based on collection of particles in sampling media, sample preparation, microscopic imagery and AI-driven partcile analysis. Give you data on pollen including species.
Pollensniffer - Collects pollen on street level; manual sampling and detection
Please leave the feedback on this idea
Shubhankar Kulkarni2 years ago
The data on the location of the specific allergens could also be saved with the date they were identified on. Some allergens may be in a higher concentration on certain days or seasons. It could be useful for a forecast if on certain days, a certain area could not be sampled since a car carrying the sensor did not go there.
Please leave the feedback on this idea
Delivery drones with integrated sensors
jnikolaMay 02, 2022
The same as cars, the drones could also be equipped with pollen detectors to provide real-time pollen data to users. This could be a feature of delivery drones mentioned here.
Please leave the feedback on this idea
Add your creative contribution
General comments
Michaela D2 years ago
Weather.com does some allergy predictions. They use Watson algorithm which combines weather data with allergy cases. By scrolling down on the daily weather prediction you find "Pollen", by clicking on it, you find a more detailed report:
This is just for pollen and I am not sure how accurate it is. Using monitors as you suggested will be more diverse and accurate at the same time. I would also like the option to combine it with an allergy symptom tracker!
Subash Chapagain Apr 14, 2022
Please leave the feedback on this idea
Please leave the feedback on this idea