Blue Sky Analytics: Data backbone of a Clean Air Economy

Blue Sky Analytics is a Big Data & AI startup led by Abhilasha Purwar (Fulbright Scholar | Yale | IIT alum) with a mission to provide geospatial data intelligence on ambient air quality (AQ) and source pollution emissions. BreeZo, our lightweight app provides historical as well as real-time contextual AQ data to helps users minimize their exposure to air pollution.

We are leveraging geospatial datasets with the ground (IoT) & drone measurements to provide pan-India AQ data with a 1 km2 pixel resolution (in 2019) & source emissions data from key stationary sources (in 2020) such as power plants, brick kilns, industrial facilities, waste dump sites, farm fires, etc to empower various public, private, and social stakeholders to lead and demand action.

By leveraging data and technology to increase transparency, accountability, and speed of action, we intend to transform the environmental monitoring and regulation system in India and globally.

India is suffering from a severe air pollution crisis, yet the awareness among the common public is dismal and policy solutions are short-termed, hollow and rarely targeted at root causes. The primary reason for this state of air, is the lack of a reasonable amount of data on key parameters about air pollution and source emissions, to even begin to take the right actions.

15 of the top 20 most polluted cities in the world are in India. Air pollution is now the second largest health risk factor. Chronic Obstructive Pulmonary Disease (COPD) is the largest killer after heart disease, responsible for 1 million deaths in 2017.

Currently, India has 150 (government-installed Beta attenuation) air quality (AQ) monitors, out of which 36 are installed merely in Delhi. This has led to uninformed & biased discourse. Example: over-focus on Delhi & resource allocation for solutions like outdoor air purifiers while not grasping the overall airshed structure and taking meaningful measures to eradicate root causes. This provides us a market & social opportunity to plug this prevalent data gap by providing AQ data and analytics at high-spatial-resolution for a fraction of conventional monitoring costs.

300+ million Indians across the entire stretch of Northern plains (from Delhi to Calcutta) breathe hazardous air for 100+ days per year. The air is laden with particulate matter (PM2.5 & PM10) 8 to 15 times the safe limit, inflicting irreversible health damages and compromising lung capacity, reducing about 4 to 10 years of lifetime.

At present, India has 150 air quality monitors covering 3.3. million km2 area with an average of 1 monitor per 600 km. Installation of one such monitor costs $250,000+ & $1000+ in monthly maintenance. This leads to sporadic data leaving even the educated masses unaware of air pollution and its harmful effects.

Globally, air pollution affects over a billion people, often in developing countries with limited addressable budgets. WHO data states that 7 million people worldwide succumb to air pollution borne ailments every year. Yet, the data on basic parameters needed to solve this problem is minimal and often lies in silos. We aim to replace bulky pdf, countless emails and piecemeal non-collaborative projects, with near real-time, high-frequency, high-resolution air quality and air pollution emissions data disseminated by modern APIs and platforms.

We have built an automated multi-source data ingestion platform for various geospatial products from ESA & NASA and on-ground monitors (government or low-cost) using AWS cloud infrastructure. Currently, all the government monitors, over 1700 low-cost monitors installed by different private, and non-governmental agencies, and data products from 4 different satellites are hooked to our system. 

The geospatial datasets are stored in S3 data lake for further cleaning. It is then pre-processed with Serverless Lambda architecture due to large memory footprint & low costs. We are one of the few firms in India to use a PostGIS for geospatial queries and time series databases (TimeScale DB) for improved processing, scalability, and resilience of our systems. Using artificial intelligence & hybrid (AI + deterministic) models, we convert these complex data-streams into actionable insights on the cloud with services such like AWS EMR & Sagemaker. We refer to all this plumbing bundled together as our geospatial data refinery which powers all our data dissemination platforms.

Through this proprietary algorithm, we will be able to provide air quality data with 10000x better spatial resolution at a fraction of cost, compared to current ground measurement methods. Even with a 15% error rate, in a country like India with PM 2.5 & PM 10 levels ranging in 250-500 ug/m3 range, we will be able to provide representative information. For the cost of installing a low-cost monitor network aided by one regulatory standard AQ monitor in one city (for example Kanpur), we can deliver AQ data for the entire state of Uttar Pradesh

On the customer side, we disseminate these actionable insights through raw data APIs, visualizations as embeds, and freemium web & mobile App BreeZo. By these channels, we are able to cater to a large spectrum of audience with varying levels of sophistication from moms to insurance analysts to scientists thereby maximizing our impact.

Our vision is to build Bloomberg for environmental indicators to enable informed decisions & actions through fast, accessible, and relevant information and analytics.

Every year, millions of people become victims of air pollution. More than 300 million people across the Indo-Gangetic plains breathe toxic air which not only impacts their health but also degrades the quality of their life.

To mitigate air pollution, public and private institutions in India carry out air quality monitoring using manual systems, Beta Attenuation Monitors, and low cost IoT devices. These are mostly installed in urban and semi-urban areas, leading to sparse data collection and unfair allocation of resources.

With BreeZo, air quality information can be obtained at a very high spatial resolution of 1km2. The app is user-centric enabling common people, technocrats, governments, and enterprises to use air quality data as needed.

The underlying technology of BreeZo is a huge geospatial data refinery with automated pipelines (AI-driven) which fetch data from satellites, government installed systems and low cost IoT monitors. The machine learning algorithms we've developed clean, organize, and process this huge corpus of big data to give intelligence on air quality which is rendered as analytics on BreeZo.

Our solution & innovation is targeted at making environmental data valuable and easily accessible. Blue Sky is building products which transform existing environment monitoring & regulation system. With the use of deep tech, we want to change the age-old mindset which perpetuates environment protection as someone else's job.

Blue Sky Analytics is a deep-tech startup engaged in big data analytics leveraging AI/ML models to fill the spatial and temporal gap in environmental data. Our proprietary geospatial data intelligence platform is built via a new and innovative application of multiple existing technologies. 

• Public cloud infrastructure by AWS to build the backbone of all our solutions
• Time-series database (TimeScale DB) to efficiently store and query IoT and environmental data
• PostGIS for storing geospatial data & queries
• Data from third-party IoT monitors to calibrate and train ML models
• Geospatial datasets like AOD, BLD, NDWI, NDVI sourced from ESA & NASA via automated pipelines
• Serverless architecture with AWS Lambda function for pre-processing the data from satellites
• AWS EMR & Sagemaker building deep learning models & analyzing the data

In addition, our innovation is the stakeholder mapping and partnership that we have built. We believe in a collaborative approach of solving complex problems by bringing various actors on the same page with optimized separation of concerns to yield the best outcome. Even as a very young startup, our deep collaborations with various academic, policy, governmental, regulatory, and civil society institutions are equivalent to the core technology innovation.

Large spatial and temporal gaps in the air quality & emissions data has made air pollution monitoring system unreliable. Air pollution is considered a Delhi problem or being caused just by vehicular emissions, while data outlines it as a North India problem caused primarily by crop burning and industrial emissions. Manual monitoring makes turnaround time for this data as 3-10 months, in some cases, 4-5 years. Blue Sky systems & technology would provide air quality and emissions data is available at daily/weekly frequency. With our innovation, we expect to change the following:

1. Transparency, & accountability: with media coverage of near real-time emissions data, regulatory agencies can initiate swifter actions against offenders. Instead of the current, unrepresented and self-reported monthly or quarterly emissions data, high-frequency & high-resolution emissions data from an independent satellite-based technology will transform the regulatory system.

2. Transformation of the APCD market: The industrial ecosystem will invest in pollution abatement equipment rather than pollution hiding measures (i.e. bribing). At present, the Air Pollution Control Device (APCD) industry is based on the existing regulatory framework which promotes installing devices but their efficacy is unchecked, leading to a proliferation of low-quality cheap devices.

3. Competitive federalism: High-resolution and high-frequency data will be made available for each industry in all states. For a power plant officer in Karnataka, emissions data for a plant in Maharashtra will be accessible, promoting cooperative federalism among states to combat the menace of air pollution together in industrialized centers of India

Our users play a vital role in app development by providing feedback and engaging with us. At present, we have more than 4400 unique users and around 5—70 people visiting BreeZo every day. Our web and mobile app recently crossed 10,000 views with an average of 2.35 mins spent per visit. In the 2019 pollution season, we are targeting to expand our base to400,000 users. In the coming five years, we want to become India's go to air pollution App with a user vase of more than million people. 

In addition, we are building various enterprise or B2B products for relevant public, private, and social stakeholders who are a part of the broader ecosystem. We aim to play a significant role in transforming and modernizing the Indian environmental monitoring and regulation system. The data and analytics is intended to enable informed decisions, catalyze swifter action, and to assist in reducing the air pollution in India by 30-50%. We expect these solutions to indirectly affect 50-60 million indirect users in the North Indian Airshed who would benefit from the clean air as an outcome of ecosystem effort. 

BreeZo, launched in Oct 2018 as Beta version, gained over 4400 unique users who logged in 10,000+ times to check air quality data spending average 2.34, all with $0 marketing budget. Since then, we have revamped the app design with a user-centric focus including various vernacular & gamification features and adding pan-India spatial coverage with our Smart AQ algorithm. For 2019, BreeZo is aiming to become India's go-to Air Pollution App and gain about 250,000 to 400,000 users.

In the next 5 years, Blue Sky Analytics aims to become one of the leading global geospatial data intelligence firms with a niche and brand identity within the environmental sector. We aim to play a significant role in transforming and modernizing the Indian environmental monitoring and regulation system. We hope to enable the entire public, private, and social ecosystem with data and actionable insights to assist in reducing the air pollution in India by 30-50%.

Our project is a complex engineering undertaking, and though we have a team of dedicated and skilled developers & data scientists, we are facing multiple technical, financial, and go-to-market challenges.

• As a young team with limited exposure to experienced managers and tech-mentors, we need help in workflow structuring and architecture.
• To deliver our product, and build the complex geospatial data refinery, we need to add 4-6 more highly skilled, motivated, and talented members.
• The heavy size of satellite products makes processing and data management a herculean engineering challenge. Next, we need to enhance our knowledge and understanding of ML /AI models to efficiently deploy the best solutions.

For the next five years, our biggest challenge will be in raising capital and building key business partnerships 

• We are operating in a relatively new tech-space, where revenue models are constantly evolving. Space-tech firms like Descartes Labs, Earthcube, and SpaceKnow have been doing good business, but an established go-to-market strategy is yet to evolve. Hence, a critical challenge is to obtain patient capital which can allow us to build technology & products while scoping and exploring business partnerships & customers.
• Next, we want to become a key data provider to the regulatory agencies, which would require strong credibility and brand recognition. Similarly, cracking partnerships in the ESG investments, respiratory healthcare, and insurance sectors are not so simple and straightforward.
• Expansion into other domains such as water quality monitoring and geographies would yield similar technical and business challenges. 

1. We are solving for the technical product development challenges by seeking out various CTOs, senior developers and solution architects, and various professors as scientific or technical advisor and allocate 3-4 hours/ week of their free time.
2. We are participating in various incubation programs to solve for customer, partnership, and go-to-market challenges.
3. We are also looking to increase our organic PR to both increase applicant pipeline for various open roles and customer pipeline for different products specially our key product BreeZo.
4. To solve for capital, we have been engaging with various foundations and impact investors for past 8 months and intend to showcase our potential and stellar candidature for investment or grants through our products.
5. Similarly, for partnerships and collaborations, we think the product speaks for itself wherein the team allocates majority of time into AI/ML modelling, data cleaning and processing, and building the dashboards.

We believe the sustained efforts and product development would establish credibility and brand recognition for us that will solve all of the above mentioned challenges.

Blue Sky Analytics has two registered entities under India's Ministry of Corporate Affairs, a Section 8 non-profit company and a private limited company under the Companies Act 2013 also recognized as a "startup" under the Department for Promotion of Industry and Internal Trade. Both entities are maintained at a arm's length distance through a master's services agreement outlining rules of their engagement. 

The purpose of Section 8 non-profit company is to ensure proper dissemination of data &  products to relevant governmental & non-profit institutions in case such products are not affordable to these institutions. While our guiding mission is building the technology and data backbone of a clean air economy, we understand that many institutions in developing countries who can best utilize and leverage these systems might be under-budgeted.

On the contrary, the private limited company focuses on value creation by building technology and intellectual property that is then provided as data, analytics, and infrastructure as a service to different stakeholders for revenue, growth, and profitability.

This hybrid structure allows us to operate as a lean deep tech startup and establish ourselves in the tech community while also ensuring maximum data dissemination and impact creation through collaboration & partnerships.

Currently, we have a 10 member team of 2 founders, 4 full-time employees, 2 interns, and 2 consultants. Abhilasha (CEO) and Kshitij (CTO) together lead the strategy, business and product development. The core team comprises of Prachi (Strategy), Jashan (Tech Infrastructure) and Saheel (Data) who are all young computer science graduates. We have two senior consultants on boards: Sheekha (Sr. Data Scientist) and Harman (Interaction Designer), to assist with hiring, building roadmap, and undertaking priority work while we add more team members for tech, data, strategy, and design roles
- Team Blue Sky

Blue Sky is a team of millennials deciding to plunge to solve one of the most critical challenges faced by Indian Society i.e. North Indian Air Pollution Crisis.

Blue Sky is led by Abhilasha (Female, 28) with global work and living experience across 5 continents i..e US, EU, Latam, West Africa, India, and SE Asia. She has worked for 3 years with India's Ministry of Environment on designing and implementing market-based Air Pollution policies. Her previous team wrote the currently in-place, guidelines for IoT sensors for industrial emissions monitoring and she played a key role in designing the data validation protocol and data acquisition and handling system. Post that, Abhilasha went to Yale as a Fulbright Scholar to pursue a master's in Environmental Management and majored in economics and sustainable finance.

While she brings the network, financial capital, environmental expertise, and a global business/strategy understanding; Kshitij is a 24-year-old college dropout and self-taught developer with a passion to build fast and efficient products using the latest tech-framework. Kshitij is plugged into the global developers' network as an open-source enthusiast and contributor. He was also able to mobilize as self-selected young, smart, and energetic team of developers and data scientists in a mission to use technology to solve environmental problems.

Together we are able to merge the two ends of the environmental and technology world and build Blue Sky at the intersection, which places us in a unique position to be able to solve this problem.

1. Knowledge partnership with CERCA, IIT-Delhi (Centre of Excellence for Research on Clean Air), and Dr. Sagnik Dey (IITD) onboard as a Scientific Advisor.

2. Incubated at NASSCOM 10000x startups for BreeZo growth, marketing, and partnerships.

3. Member of T-Hub Facebook India Innovation Accelerator for AI mentorship & technology support

4. Resident at The Common Room Summer Residency focused on system design mentorship

We are working on small pilot project with CPCB (Central Pollution Control Board), and are in conversations with EPIC, Shakti Foundation and Azim Premji Philanthropic Initiatives for collaborations.

Abhilasha (CEO) represented as a North India Air Pollution lead in the Kathmandu TechCamp: a U.S. Department of State initiative. In addition, we are exploring data partnerships and API integration with many private enterprises in mobile, healthcare, and insurance sector; that will be announced as they materialize.

Blue Sky is a freemium Data & Analytics as a Service (DaaS) business with three categories of clients:

1. B2B2C eyeballs based revenue via BreeZo mobile app, Ads & media visualization sales,
2. B2G subscription of BreeZo web platforms by various government and regulatory agencies, and
3. B2B subscription of BreeZo APIs and dashboards for improved analytics on various products by sports apps, healthcare, and insurance industry.

We capture value by delivering relevant data and insights which enable the customer to make direct or indirect improved decisions. Our key resources are our team, scientific advisors, angel investors, and the global community which has a growing concern regarding air pollution and broader environmental crisis. As a software product, initially, we will have a direct sales channel with relevant partners & stakeholders. Eventually, we want to build deep integration with key data providers like Google or iOS, to disseminate data through rich snippets on a google search or native banners on iPhone home screens.

As a young startup, we are building our key partnerships; currently talking with Xiaomi (the leading mobile seller) to place BreeZo as a native app on over 250,000 phones. Similarly, we are talking to senior ranking officials in cities like Kanpur, Ranchi, and Indore, which do not have a single installed monitor and rely on poorly calibrated low-cost monitors for air quality data, to provide them subscription-based satellite-derived air quality data

In the coming decade, all the parts of the world will be faced with various types and levels of environmental challenge and will need to swiftly and efficiently make sustainable, environment-friendly decisions and systems. 

Whether it is crop burning in Punjab, water crisis in South Africa or the heat crisis in Middle-East; public institutions, private companies, insurance providers, investors all would increasingly require access to better environmental data to understand their own impact, take decisions regarding various projects, quantify risks in the warmed new world, or proactively lead a policy change. 

Governments, foundations, global universities, and public institutions are becoming large consumers of such data. Thus we believe that there is a large white space and an exponentially rising demand in the environmental data market. 

On the investment side, we are seeking financial stability via an innovative and hybrid mix of equity capital for technology development and grants for data and analytics dissemination to any non-profitable categories. From a sustainable business perspective, the global geospatial analytics market is expected to reach $175 B by 2027 with a CAGR of 18%, Earth Observation (EO) market is pegged at $7 B annually by 2027, and global air pollution monitoring is expected to reach $5.5 B by 2024. 

Leveraging AI and geospatial data, we aim to become a global provider of environmental data as a scalable profitable business. 

As a satellite data analytics startup focused on environmental parameters, MIT Solve - world's leading incubator for tech entrepreneurs solving global issues is a perfect partner to us. With the help of Solve, we expect to leverage the mentors & other startup community for technical mentorship, refining business model and gaining global access.

Our project at Blue Sky Analytics is a very complex engineering undertaking, and though we have a stellar team of very dedicated, and experienced developers & data scientists, we still need abundant help to solve our technical challenges. We are in a fairly nascent global market with yet-evolving business models & use-cases, where guidance from the Solve team can be principal to our growth. Efficient management of our huge data corpus and improvising data processes while using our financial resources effectively is essential for us to deliver more value to the end-users. Assistance from Solve experts in managing our data can be critical in achieving this end.

Environmental issues and data have, over years, transformed into the boring, don't care category at various levels of Indian public, governmental, or private engagement. The core of disseminating data is the design of the web or mobile medium; which intrigues and demands the users to "care". Becoming a part of the Solve community would make BreeZo a brand and also change the narrative around conventional environmentalism.

The equity-free grant offered by Solve would help us allocate capital resources in design, development, marketing and dissemination of better data products and platforms.

We need more technology mentorship in big data analytics, AI and Machine learning. We are a young and enthusiastic team of passionate environmentalist developers and data scientists, but we need more mentorship and guidance to build best solutions in the most optimized manner.

If possible, we would like mentorship and help from MIT Media Labs in designing various technology platforms, mobile and web Apps, and dashboards for different users. 

• Public institutions are the most significant stakeholders in the environmental ecosystem. We must build deep partnerships with various environmental ministries, regulatory agencies like National Green Tribunal and Courts in India, and monitoring agencies like state and central pollution control boards 
• Various municipal corporations and leaders in local or state governance who frame important environmental policies or undertake major projects. 
• Google, Facebook, Xiaomi, and Bloomberg for optimal and maximum data dissemination. AWS and TimeScale DB for product development and data management. 
• Partnerships with leading print, digital, and broadcasting media agencies to enable the highest data and visualization dissemination. 
• Blue Sky sources satellite data from NASA, European Space Agency and this knowledge and data collaborations with them would be highly significant. We would also like to partner with ISRO to leverage the potential of remote sensing data. Further, we would like to build partnerships with private data providers like Digital Globe and Planet for select high-resolution raw imagery. 
• API integrations with respiratory healthcare companies like GSK, Novartis, Cipla, etc to integrate air quality data in deep analytics to improve supply chain and digital marketing strategies.
• API integration with life and health insurance industry to integrate air quality as a pricing factor to catalyze a free-market based environmental awareness and call for action. 
• Engage with think tanks, advocacy groups, and civil society organizations like UN, WRI, ORF, Gates Foundation etc, to spread awareness about air pollution, emissions, and its impact on human health and environment

AL and ML form the base of our technology innovation used to predict pan-India ambient air quality and estimate source emissions from various point and area sources like brick kilns, coal power plants, farm fires, waste-burning sites etc. In future, we want to use similar system to do surface water and effluent monitoring across various rivers & lakes in the country.

If selected for the AI Innovations Prize, we would like to utilise the prize in achieving the following targets by next year -

• Expand our data and tech team by hiring best technical talent, build strong AI/ML capabilities and and explore more ML models to process geospatial data for different parameters
• Do more holistic user experience research and invest in design and development of BreeZo Air Pollution App. 
• Use some capital for growth and marketing of BreeZo, our web & mobile Air Pollution App and invest in features like gamification and adding different vernacular languages, to reach maximum user base.
• Training via paid workshop & upskilling of the team like registering our team members for MOOCs and short duration online course 

Air pollution affects over a billion people globally, all strata of society alike, in particular, lower-income groups in inferior neighborhoods (next to industrial units, waste sites, or highways) are more exposed to the pollution. By increased transparency and data dissemination, Blue Sky aims to reduce airshed level air pollution by 20-30% in India. 
If awarded the Innovating Together for Healthy Cities Prize, Team Blue Sky would utilize the prize to scale the business and do focused work on the following -

• Fostering awareness about air pollution and its impact on human health.
• Partner with schools, colleges, and universities in data dissemination and using BreeZo predictions to declare holidays on critical air days. 
• Develop an air pollution game to make air pollution concepts easy to understand for all types of audiences. 
• Invest in PR and marketing of our app BreeZo using all forms of print and digital media, like Facebook and Instagram ads
• Build the platform to monitor emissions from industrial clusters, waste sites, and farm fires for improved policy implementation and legal enforcement. 
• Air Pollution is the second largest killer in India, affecting millions of low-income bike-riders who face the highest exposure. BreeZo aims to make them more aware, provide them with the best solutions (i.e. masks) to protect themselves, and lead the change in overall regulation system to improve air quality directly affecting health and lives of millions. 

Not Applicable

Indoor air pollution is a gender issue. 43% of the global population still uses coal and its derivatives for cooking and heating leading to pulmonary diseases, asthma, and in many cases premature death. Air pollution is also a menace for pregnant women and lactating mothers.

If awarded with the Innovation for Women Prize, we will use this prize for -

• bringing awareness about household air pollution, its effect and alternate means of cooking and heating by organising workshops and developing easy to understand content
• develop data solutions specific to this concern and partner with relevant public institutions, healthcare institutions, and women's organisations
• invest in PR and marketing to drive the public discourse around household air pollution. Make infographics, short explanatory videos and graphics about household air pollution. 
• Lead movement for overall pollution reduction to directly benefit young females, pregnant and lactating mothers. 

Blue Sky Analytics is a female-led deep-tech startup building AI/ML-based solutions for environmental problems like air and water crisis which disproportionately affect women and place them at higher exposure and risk. 

Not Applicable. Though our solution utilized data at the core of our innovation, we are not a US based entity and are presently not focused on America as a geography. 

Not Applicable.