Folia

According to a study conducted by the U.S. Center for Disease Control, farm workers are up to 20 times more likely to suffer from heat stress and heat-related illnesses, such as skin cancer and heat stroke. In fact, nearly 40 farm workers every year die from heat stress in outdoor farming. Of these casualties, the most vulnerable populations are seniors (65+).

Why is this relevant?

Most of the U.S.’s farmland is located in non-coastal, inland states. These inland states also happen to be disproportionately affected by global warming, reaching record-high temperatures with low humidity in recent years. Thus, as climate change persists, it is becoming more and more dangerous to be exposed to the sun’s UV rays for long periods of time. And what greater victims of this are there than outdoor workers?

Part of the work of a farmer, particularly in large monoculture farms, includes maintaining a watchful eye and removing diseased plants as quickly as possible. This is because in monocultures, there’s limited genetic diversity, so it’s very easy for foliar diseases to spread rapidly, potentially costing thousands in crop loss and exacerbating the issue of food insecurity. However, the current method of preventing this requires workers to spend long hours sorting plants by hand which, as aforementioned, increases the health risk. Even when not sorted by hand, operating heavy machinery requires one to spend long hours outdoors, which can still take a toll long-term and can make already vulnerable populations more susceptible to heat-related illnesses.

Folia, a Python-based deep learning algorithm loaded onto a Raspberry Pi camera (held together by a 3D-printed Polylactic Acid chassis) that utilizes feature extraction to detect the beginnings of common rust, gray leaf spot, and northern leaf blight on plant leaves. Once detected, through a bot, the system sends an SMS message to the user’s phone, alerting them of the type of disease and the coordinates of the specific plant (also known as geotagging).

How would it work?

• The algorithm itself is run on OpenCV and would use object detection from the camera to consistently monitor crops within a 50-square-foot radius and it’d tested and trained specifically for whatever crop is being grown in the mono-crop.

• As the plants grow, once the spores of common rust, gray leaf, or northern leaf blight start to show on the plant leaves and within view of the device, the remote device will send the alert to the user.

Folia’s target population is mainly middle-aged adults (40-60 years old) and seniors (60+) who live in the United States and own any amount of farmland, be it for commercial or personal use. It would also be highly suitable for individuals with medical conditions that make them more susceptible to heat-related illnesses and individuals with a personal/family history of contracting heat-related illnesses. 

Folia intends to improve crop management in a way that mitigates health risks for the farmer. This includes offering a solution that limits the need for extended outdoor exposure by eliminating the issue of having to monitor crops manually. It also decreases the risk of contracting food-borne illnesses, which can occur if fungal diseases in crops go unnoticed and therefore, aren’t pruned.

Folia was initially created to target the issues of the people within its own community. The state of Maryland has over 6,000 full-time farmers, 2 million acres of farmland, and agriculture as its largest commercial industry. Maryland is also home to various federal research facilities for the U.S. Department of Agriculture (USDA). As such, concerns for the health and well-being of Maryland farmers are a growing concern within the community. In fact, Southern Maryland Online reports that 78% of Marylanders feel the need to support Maryland farmers.

Folia endeavors to combat this issue by factoring in the insights of its home community and applying them on a broader scale. We are already in discussions with local USDA facilities in relation to finalizing the prototype, testing, and deploying Folia.

Folia was developed with 3 goals in mind:

• Devise a strategy for leveraging remote sensing technology to reduce the prevalence of foliar diseases by detecting them faster than a human would. This way, diseases are less likely to spread, and fewer crops will be lost.

• Implement this strategy in a manner that effectively mitigates the human health hazards inherent to prolonged outdoor agricultural engagement, such as heat-related ailments.

• Execute these endeavors with a focus on a cost-effective design with biodegradable materials.

Folia is innovative in its health-related implementation of remote sensing technology. Governments and well-established companies often use remote sensing technology for crop monitoring, but not typically for health preservation. 

Folia is also unique in its use of remote sensing technology in a manner that's not only cost-effective but also environmentally sustainable. Given that unsustainable raw materials only exacerbate issues of climate change and its health-related side effects, we took the initiative of developing as sustainable a technology as possible. The device's exterior is fully biodegradable, and the camera contains non-plastic, recyclable materials. The device also only costs $500 to make in terms of raw materials alone, which is considerably below the market average. This allows it to be easily implemented and distributed to those who need it, as we don't believe in money being a barrier to health and well-being.

Folia promotes UN SDG 3 by acting as an early warning system that reduces contamination, reduces health risks, and tackles premature and preventable mortality. More specifically, it addresses 3 of the target areas of SDG 3:

• 3.9: By 2030, substantially reduce the number of deaths and illnesses from hazardous chemicals and air, water and soil pollution and contamination. 

Foliar diseases also present the risk of groundwater and soil contamination, according to the National Institute of Health (NIH). The bacteria that cause foliar diseases in plants can inadvertently lead to water contamination through harmful bacteria being swept up by rainfall and contaminating groundwater sources.

• 3.4: By 2030, reduce by one-third premature mortality from non-communicable diseases through prevention and treatment and promote mental health and well-being. 

Most heat-related illnesses, such as heat stroke and Rhabdomyolysis (muscle breakdown), are non-communicable and yet can often be fatal. By mitigating risk, we are also limiting the death toll of curable diseases.

• 3.d: Strengthen the capacity of all countries, in particular developing countries, for early warning, risk reduction, and management of national and global health risks.
  

With ambitions of expanding nationally and intercontinental, we hope to implement Folia in underserved regions (namely in the global south, where the majority of agricultural exports originate) to address even more dire instances of preventable mortality through unsustainable agricultural practices. Especially abroad, Folia's risk reduction technology could save millions of lives.

Data Acquisition

The data on which Folia was trained come from a mix of plant leaf image datasets that were manually collected through field research, as well as publically available datasets from the PlantVillage project, courtesy of Pennsylvania State University. As we continue to refine our algorithm, we expect to continue gathering visual data from local farms, while creating a varied training set with mixes of photos from government agricultural research facilities and private farms.

Software Development

The AI technology that powers Folia was developed using OpenCV in Python. More specifically, ORB (Orientated FAST (Features from Accelerated Segment Test) and Robust BRIEF (Binary Robust Independent Elementary Features)) and SURF (Speeded-Up Robust Features) to extract features from all reference images (datasets) before cross-referencing them with each frame of the device camera, which takes pictures at a frequency of 2 frames per minute currently. 

Consent and AI Ethics

At Folia Technologies, as a humanitarian non-profit, we pride ourselves on prioritizing ethical considerations during technological development and data acquisition. This is why we ensure that all parties involved in our data acquisition process (Penn State University and various local farms) provide complete consent before utilizing their data to train models. We also provide our data sources anonymity unless they suggest otherwise.

Risk Mitigation and Management

We employ standard cybersecurity tools in order to mitigate risk, including encrypting data in databases, utilizing an on-boarding process for staff/volunteers that includes thorough security and protocol training, and restricting sensitive information to authorized personnel.

3

1.5 years

As a black-owned and women-led organization, Folia Technologies is very committed to diversity and inclusion in the workplace. This is why our bylaws contain clauses that promote equitable recruitment among both employees and volunteers, including necessitating at least 30% of employees be female/gender minorities, at least 50% of employees be BiPOC, and at least 20% of employees be from overburdened and underserved communities (as highlighted by the U.S. Climate and Economic Justice Screening Tool).

Our current operating costs are $27,400. This includes human capital, product design, business operations, materialand research and developmental costs. Our operating costs for next year are projected to be about the same at $25,000.

To fund our endeavors, we are seeking $70,000 in funding to cover outreach, product development, staffing, and other operations. 

The Cure Residency would be an incredible opportunity for our team. The opportunity to receive personalized mentorship from interdisciplinary health professionals is unmatched, as well as the lucrative educational and networking opportunities for professionals and like-minded social innovators in Health AI.