ENGINEERING human dignity
Cinema has the power to change how we see the world. For me, a Bollywood movie "Article 15" didn't just tell a story; it exposed the invisible plight of manual scavengers in our society. That wake-up call led me to develop my capstone project .I’m Arnav, and I believe technology should serve those whom society has overlooked.
CONSTITUTIONAL MANDATE
Article 15 of the Indian Constitution strictly prohibits discrimination on the basis of caste; yet, the persistence of manual scavenging reveals a deep-seated violation of this equality.
ARTICLE 15
.png)
"Each of us owe it to this large segment of our population, who have remained unseen, unheard and muted, in bondage, systematically trapped in inhumane conditions."
-Justice S. Ravindra Bhat
15
Introducing
CLEAR FLOW
This is my bid for my society, introducing clear flow a patented manual scavenging machine used to transform manual scavenging into automated scavenging
To eliminate human entry, we have designed the Clear-flow, a specialized Robotic platform engineered to operate within the high-toxicity, high-humidity, and unstructured terrain of subterranean sewers. The chassis utilizes a Hybrid Locomotion System combining high-torque treads with articulating stabilizers to navigate through viscous sludge, varying pipe diameters, and vertical offsets.
RESEACH BEHIND
CLEAR FLOW
Clear flow is backed by systematic research with a clear scientific methodology , explore our entire research paper at enodo. The Dataset behind the research is hosted at Harvard Dataverse
I have tried to map the "flow" of our cities, not just to make them more efficient, but to make them more humane. If our algorithms can predict a crisis 30 days in advance, then every day we allow a manual scavenger to enter a manhole is a failure of our collective imagination, not our technology.
To effectively model the spatial dynamics, we define the sewer network as a weighted directed graph where the topology is dictated by the gravity-driven flow of wastewater.
The nodes in our graph represent manholes, junctions, or pump stations, while the edges represent the sewer pipes. Unlike traditional road networks, the sewer network is a directed acyclic graph (DAG) in most sections, though loops may exist in combined systems.
We merge Static Metadata including pipe material, diameter, and hydraulic gradient with Dynamic Maintenance Logs, which are transformed into a numerical Failure Intensity Index. To account for external catalysts, we integrate Exogenous Stressors such as cumulative precipitation and NDVI based vegetation mapping to track root intrusion risks.
To facilitate deep learning, the synthesized features are structured into a Spatio-Temporal Tensor, a three-dimensional array representing Nodes, Time-Steps, and Feature Channels. Given that municipal records are often sparse or irregular, we employ Zero-Infill Imputation and Temporal Masking to maintain a continuous data stream without introducing artificial bias.
गरिमा
.png)