Aquasaic

In response to the Global Climate Challenge, our project targets a critical and escalating issue exacerbated by climate change: the increasing pollution of freshwater sources, with a specific focus on the situation in Antioch, California. This local example highlights a global crisis where the adverse effects of climate change—such as droughts and sea level rise—combine with human activities, including extensive water distribution and environmental mismanagement, to jeopardize water quality and availability, ultimately affecting the Blue Economy .

Globally, climate change is intensifying natural disasters like droughts and floods, leading to significant environmental and socio-economic impacts. These include loss of marine biodiversity, damage to marine-based infrastructure, and the degradation of coastal environments. For communities invested in the Blue Economy, the repercussions are profound, manifesting as economic challenges through damages to fisheries, increased operational costs, loss of tourism revenue, and the necessity for costly water purification efforts. 

For other communities, such as those in Antioch, California, climate change is posing a direct livelihood challenge. Situated on the banks of the Delta, Antioch has historically relied on this freshwater source to support its population of over 110,000 residents, many of which live in Environmental Justice (EJ) neighborhoods. Recent historic droughts in California, a direct consequence of climate change, have necessitated the redistribution of local freshwater to meet the demands of broader Southern and Central California. This redistribution, coupled with sea level rise and decreased rainfall, has led to a steady increase in the salinity of Antioch’s water supply, hindering the city’s ability to provide clean and safe drinking water. The situation has escalated to the point where a $110 million desalination plant is deemed necessary and is expected to be operational by 2024 to address the issue. 

To our knowledge, water in Antioch is not being treated to accommodate the local environment it houses, and the upcoming purification efforts are currently focused on purification-for-distribution, not to accommodate the pre-existing marine life that allows waters to be environmentally resilient long-term. 

This local scenario mirrors a global crisis affecting millions, underscoring the urgency of addressing the multifaceted impacts of climate change on water resources. The challenge we tackle through our initiative is not isolated to Antioch but is emblematic of a widespread issue confronting communities worldwide. By implementing sustainable and innovative solutions to combat water pollution, we aim to mitigate one of the many detrimental effects of climate change.

Our focus on Antioch’s water salinity issue highlights the necessity for concerted efforts to combat climate change and its impacts on our planet’s water resources. Through our project, we contribute to the global dialogue and action towards a more sustainable and resilient future, addressing the urgent need to adapt and respond to the environmental challenges posed by our changing climate. Ultimately, our goal is to enter the water remediation dialogues, and provide a service that improves the methodology to remediate water, allowing companies, local governments, and policymakers to make holistic decisions that concern water.

We integrate (i) physical surface data from NASA’s EARTH satellite, (ii) chemical information from the U.S. Geological Survey, (iii) biological data from the National Center for Biotechnology Innovation (NCBI)'s databases, and (iv) an in-house-developed techno-economic analyses of available water remediation processes, using a computational pipeline. Our Large Language Model (LLM) Machine Learning algorithms then are capable of (a) assessing water resiliency, as indicated by native organismal survivability, and (b) determining a cost-efficient and sustainable water remediation process, which will result in safe water bodies, capable of proliferating local, environment-beneficial organisms. 

First, our customer engages with us by providing us with a representative water sample. Next, our team characterizes the water physically (temperature, turbidity, total suspended solids, total dissolved solids), chemically (pH levels, nutrients, pesticides, volatile organic compounds, oil and grease, endocrine-disrupting chemicals, pharmaceuticals, radioactive materials), and biologically (RNA traces of bacteria, fungi, and viruses). Physical and chemical analysis will be performed by our analytical chemists using spectrophotometers, chromatographs, colorimeters, pH meters, and mass spectrometers. Biological datasets will be analyzed by our microbiologists. Lastly, our computational team integrates this data, and analyzes it using our custom, machine learning-based pipeline described above. We envision this pipeline to ultimately land in our customer’s online dashboard, allowing them to make holistic decisions about water handling that will result in increased water quality resiliency.

We have begun our journey by learning about one of the coastal communities surrounding the Aquasaic team: Antioch, CA. The situation in Antioch, in which the local government has made quick decisions to remedy their water salinization challenge, has demonstrated to us how helpful we will be in acute scenarios, where governments are investing millions of dollars in new strategies to remediate water. We have also been in touch with the local governments of Cape Cod, MA, Martha’s Vineyard, MA, and Flint, MI. All have expressed concern over the quality in their bordering waters, and understand that their communities are likely to be impacted by Climate Change over the next 100 years. 

We expect our solution will serve these communities first. However, we have the potential to expand to other coastal regions, where the acute impact of climate change and local water contamination is unexpected and severe. Through our services, local governments will be able to select water remediation pipelines that will turn their communities into strong, resilient water reservoir communities. With a treatment pipeline compatible with the environment they reside in, these communities’ recreational and entrepreneurial use of their local waters will be restored, and potentiated. 

Shalmalee spent a lot of time in India with extended family and remembers being flabbergasted when she wasn’t allowed to drink the same water as her relatives. Her water was filtered and boiled, which introduced her to the concept of environmental inequity at an early age. Growing up in California where environmental consciousness is ingrained in the social milieu, she took seemingly essential luxuries (e.g. waste management infrastructure) for granted. Upon moving to Boston for graduate school, she noticed that Massachusetts lacked infrastructure for industrial composting. This highlighted the various forms in which environmental inequity exists. Inspired by the potential for engineering biological solutions to address environmental injustices, she dedicated her PhD towards the fight against fossil fuels, which she investigated for commercialization potential. She determined that this idea was not scalable, which forced her to pivot her efforts. Thinking of her relatives inspired her to evolve her technology to tackle another problem: the creation of a cost-effective scalable process to clean water.

Nikita grew up on the Red Sea coast in Saudi Arabia, where the increasingly detrimental effects of industrial waste in her childhood home on air quality, soil type, and water quality substantially changed her day to day experiences over time. Over the years, the relentless pollution caused corals to endlessly wash up on shore while severely harming human health. Her mother and sibling developed location-specific asthma symptoms that have since improved upon their relocation. As these consequences affected her family’s, her friends’, her communities’ day-to-day lives, she set out to reverse the damage done to human health and the environment. Her exploration of avenues to mitigate these effects has led to her current path of climate entrepreneurship, where we are efficiently targeting these challenges. Leveraging her bioengineering background to explore aging in the body, she realized the depth of impact food and its sources have on our physiology, and my work with Project Xylem inspired her to think about the lasting repercussions on individual and population health of social inequity of food and basic utilities. While academic research excels at identifying the specific bases for disease, the societal context that leads to successful implementation diverges from basic science.

Juan grew up in Northern Mexico, in a region where tap water was unsafe to drink. His family would spend a monthly allowance on the purchase of 20L water jugs. In his community, it was not uncommon for people to choose to drink Coca Cola over water. In San Cristobal de las Casas, a southern town in Mexico, water scarcity has become so severe that its population drinks Coke at a higher rate than water, causing the community to experience high rates of diabetes. Water scarcity in Mexico is a worsening problem, owed to increased temperatures and decreased rainfall. Juan believes in Aquasaic’s potential to help communities around the world, who deserve safe, drinkable water. Juan is finishing his PhD in Biomedical Engineering, and he is currently developing point-of-care biosensors to help communities test their water against emerging contaminants.

We have performed extensive customer discovery, focusing especially on cases where existing solutions are challenging to deploy. We discovered that in many of these cases, the profile of the water (geographical location, pH, contaminant characterization, etc.) renders the existing green alternatives either inefficient or impractical. From customer discovery, there is an apparent need to characterize water profiles and match them to an optimal treatment strategy:

“We need to understand the makeup and impact of water treatment strategies and policies

before enacting them.” - Government Official from Flint, MI.

“We often have complex contaminated waters that the government ends up paying a lot to

remediate.” - Project Manager, California Environmental Protection Agency.

“There are huge needs in the water management and water analytics spaces. Water is a scarce and precious resources and the industries which is touches are profound – manufacturing, agriculture, food and beverage, etc.”” – Mechanical Engineer who has worked in Mining, Manufacturing, and Venture Capital 

While we are quoting just three of our interviewees, several others echoed a similar sentiment. This customer learning has led us to develop our novel concept: at Aquasaic, we are using advanced machine learning (ML) techniques to analyze and understand water’s physical, chemical, and biological qualities in greater depths than previously possible. In turn, this allows us to provide to our customers a thorough characterization and understanding of the water they must treat, informing them of the costs and benefits of different treatment strategies based on water profile. 

We have created a prototype dashboard using Figma, a collaborative interface design tool. In it, we have outlined several metrics that our pipeline will deliver to our customer, and we are garnering interest and feedback from potential clients.

Aquasaic is applying to Solve to find solutions in technical- and business-related issues.

First, our team has a strong understanding of the physical, chemical, and biological aspects of our solution. However, we seek experienced contributors to determine the key techno-economic factors that make one water treatment solution more efficient over another. 

Second, our team is looking to tap into MIT’s community of software developers to fully realize our concept.

The complexity of biological data has created numerous challenges in mapping microbial communities to ecological data. Consequently, biological components are completely overlooked while informing decisions around water. Therefore, our innovation lies in the use of LLMs to create microorganism-powered water resiliency. 

In addition to the difficulties in understanding biological data, reconciling the data from living organisms to the data of their nonliving components (i.e., the environment they live in) can be quite challenging, as biological data can simultaneously have low and high amounts of noise. As each point of biological data is capturing a snapshot in the life of the organism, it accurately represents that organism with low amounts of noise. Meanwhile, gathering data from multiple cells of a single-celled organism leads to an aggregate (average) state of that single-celled organism, creating a high amount of noise. Advances in machine learning (ML) techniques like large language models (LLMs) have created opportunities to reconcile data from the living and nonliving to produce meaningful insights. Unique characteristics of LLMs including zero-shot learning allow for predictions even without observing related examples during training. 

Our solution provides data-driven water remediation recommendations to government agencies and water treatment facilities. 

In turn, governments and organizations can choose an efficient and sustainable water treatment process. 

Finally, the communities they serve can enjoy safe drinking water quickly and consistently. 

We will use our machine learning (ML) pipeline to determine the best remediation process for contaminated water bodies, thereby tackling water-related United Nations Sustainable Development Goals (SDGs), particularly Goal 6 (Clean Water and Sanitation) and Goal 13 (Climate Action).

Goal 6: Clean water and sanitation

Our aim is to significantly enhance water quality in contaminated areas by deploying an ML-driven remediation process. This tailored approach ensures efficient and effective restoration of water bodies, making clean water accessible to communities in need. Our measurement metrics for this goal include: 1. Number of Homes with Clean Running Tap Water: Progress is measured by the increase in homes with access to clean water in targeted areas, directly showing the impact of our remediation efforts. 2. H’ Diversity of Species: Improvement in the biodiversity of aquatic life, assessed through the Shannon Diversity Index, will indicate a healthier ecosystem post-remediation, evidencing effective sanitation practices.

We will engage with the communities we serve in order to monitor these metrics. 

Goal 13: Limit and adapt to climate change

Our initiative ultimately allows water bodies to foster the growth of synergistic microbiological communities in the treated water. These in turn allow the water bodies they live in to remain resilient, long-term, against climate change-induced acute challenges. Through our pipeline, we will be able to determine projections of water cleanliness over a period of time, which we can monitor. Our progress towards Goal 13 will be measured by comparing the actual water profile over time against our prediction. 

We leverage biology’s intrinsic healing properties. Here are two surprising ways in which biological organisms are helpful for water resilience purposes:

1. Contaminated groundwater from the Oak Ridge Integrated Field Research Center showed microbial populations with elevated genes encoding heavy metal metabolism, despite the lack of heavy metal concentrations in the environment. This suggests that the microbial populations contribute to the remediation of water by capturing the heavy metals. 

2. Another study found a nitrogenase paralog between the gut-adapted bacteria Elusimicrobia and the groundwater associated lineage, suggesting groundwater bacteria can turn nitrogen into ammonia, contributing to the nitrogen fixation cycle.

These are two examples which inspire us to believe in a future in which biology is leveraged to clean water. 

Multi-omic information, which gives us insight into certain biological agents’ ability for water remediation, can be found within extensive databases, such as NIH’s National Center for Biotechnology Information (NCBI). Through machine learning algorithms, we expect to couple this information with geographic, physical, and chemical data to identify potential sources of bioremediation-inducing microorganisms. 

We understand that water remediation can be performed using biological organisms:

Gifford S, Dunstan RH, O'Connor W, Koller CE, MacFarlane GR. Aquatic zooremediation: deploying animals to remediate contaminated aquatic environments. Trends Biotechnol. 2007;25(2):60-65. doi:10.1016/j.tibtech.2006.12.002

Moreover, we understand there is a wealth of datasets available for us to mine:

- Satellite data for water quality (https://www.earthdata.nasa.gov...)

- National geologic survey data (https://maps.waterdata.usgs.go...)

- Biological datasets that describe organisms' potential to remediate water and their corresponding, ecological homes (https://www.ncbi.nlm.nih.gov/)

Lastly, we understand that biology can be analyzed using machine learning algorithms and know we can use Large Language Models (LLM) to extract relevant ontological information from these:

- Greener, J.G., Kandathil, S.M., Moffat, L. et al. A guide to machine learning for biologists. Nat Rev Mol Cell Biol 23, 40–55 (2022). https://doi-org.ezproxyberklee.flo.org/10.1038/s41580...

- Nashwan AJ, AbuJaber AA. Harnessing the Power of Large Language Models (LLMs) for Electronic Health Records (EHRs) Optimization. Cureus. 2023 Jul 29;15(7):e42634. doi: 10.7759/cureus.42634. PMID: 37644945; PMCID: PMC10461074.

- Matentzoglu, Nicolas, et al. "MapperGPT: Large Language Models for Linking and Mapping Entities." arXiv preprint arXiv:2310.03666 (2023).

Taken together, we strongly believe we are capable of carrying out our company's vision. 

4

1 year

Diversity in all dimensions will be recruited for. We will partner with undergraduate and graduate student organizations that aim to help underrepresented students meet their potential. We will offer internships if and as funding allows. We will attend conferences that are aimed at helping underrepresented groups reach the same opportunities.

Additionally, due to the nature of our work, we will be involved heavily in community engagement to both create and learn from consensus as well as buy in for deploying our solution. Industrial pollution tends to be worse in communities that have dense populations of people of color, such as Flint, MI, and Jackson, MI, due to lack of resources. As people of color face these problems on a day-to-day basis, we believe that we cannot move forward in a thoughtful manner without building a team that accurately represents the demographic we are serving.

We will leverage the networks that we build through community engagement to not only ensure that we are building the most diverse team possible, but that diverse team is in turn serving diverse communities.

Mentorship:

We will develop a mentorship program for all team members and will especially support individuals from underrepresented populations in the way that they want to be supported. Team members will be paired with mentors to set and realize personal and professional development goals. Training and support will be given to mentors on how to effectively manage and mentor.

We acknowledge that respectful, clear, and succinct communication is not only invaluable when we are speaking with collaborators, but also in internal communication. To ensure communication and to have an open space to build a supportive, respectful, and compassionate work culture, team members will meet at least once monthly over Zoom to discuss culture building.

A certain allowable reimbursable limit or stipend will be provided for funding professional development activities. Goals for each team member will be generated and confirmed at the beginning of each quarter and evaluated at the end of the quarter. Personal and professional development goals will be included in the key performance indicators for measuring success.

Safe and Inclusive Work Environment

DEI training and unconscious bias training will be mandatory for each employee. A process for reporting any transgressions will be created in a way that does not negatively affect the one who reports the transgression. Discrimination and harassment will lead to dismissal. Additionally, a DEI committee will be created upon reaching critical mass (~5 employees) at the company. This committee will develop, implement, and oversee policies and procedures that promote and maintain an inclusive environment.

We intend to make all published materials accessible to people with disabilities through using accessible formats and will provide alternative versions when requested. Any accommodations for employees, including but not limited to screen readers, sign language interpretation, alternative workspace or work tools, and accessible transportation will be made. We will seek out office space that is accessible and hold events that are in accessible venues and can provide simultaneous captioning and/or transcription.

Our business model is focused on partnering with local governments, such as city and municipal governments, to serve as their official service partner. We can offer them insights into their water, which they can use to prepare health risk assessments and as a policy-informing tool. The health risk assessments we help prepare can be provided to citizens, enhancing public health and environmental safety. 

Our insights can also be utilized by NGOs and semi-governmental organizations for activism and lobbying efforts, extending our platform’s reach and impact beyond just governmental use.

Currently, cities like Antioch, Flint, and Martha’s Vineyard do not have access to a system like ours, whereas Cape Cod already employs a water tracking tool but could benefit from the addition of biological metrics. Our system is designed to fill this gap by providing comprehensive insights into water quality and related health risks.

We propose an annual subscription model, which is scalable based on the number of users. Each user, whether a public health official, an environmental official, or a water municipal official, can create a personalized profile tailored to their specific needs and responsibilities. Additionally, should a government choose to integrate a citizen portal into their subscription, this feature would be available for an additional fee.

Our pricing strategy is designed to be budget-friendly, with initial pricing set between $5,000 and $10,000. This price range is strategically chosen to fit into city budgets that typically allow for fast turnover without the need for extensive approval processes. This ensures that our tool can be quickly implemented and begin providing value immediately.

By adopting our service, local governments and organizations can enhance their capacity to manage public health risks and environmental challenges effectively, ensuring safer, healthier communities.

We are applying to non-dilutive funding at present, to build out our technology, demonstrate competitive usefulness, and attract our first customers. The initial R&D expenses will be covered through cash prizes. We have secured funding through Accel at Greentown Labs and will continue to seek these prizes over the next year. We are also actively replying to grant solicitations, such as DoE SBIR, MassChallenge, and VentureForClimateTech. 

As we meet our goals, we will begin raising pre-seed capital that will allow us to develop a robust, deployable computational pipeline to a larger customer base.