Somali Chaterji
Purdue University, IN·Computer Science·Early-career
Modern farms are full of cheap sensors and drones that produce huge streams of data, but turning that data into quick, dependable decisions is hard. Sensors fail or send noisy readings, networks drop out, and small devices have little computing power.
- Fields
Computer Science
Engineering
Plant Science & Agriculture
- Topics
cyber-physical systems
Internet of Things
IoT
- Grants
- 1 known grant, through Jan 2027
What they work on
- Develop on-device machine learning analytics for farm sensors that adapt to each device's computing power, changing network conditions and competition from co-located applications.
- Design approximate computation methods that run heavy streaming analytics, such as computer vision, across networks of unreliable, noisy agricultural sensor nodes while staying dependable and fast.
- Build a system that shifts computation among sensors, edge devices and the cloud, and extend it to serverless architectures and energy-efficient drone swarms for farm sensing and actuation.
More about the research
The challenge is building computing systems that still give farmers timely, reliable answers under these messy real-world conditions.
Why it matters
Reliable, fast analysis of farm sensor and drone data could help farmers and digital agriculture providers act promptly, supporting more sustainable farming. Systems that tolerate missing data, failing sensors and weak networks could make precision agriculture practical on real farms, and industry partners may help turn the methods into usable prototypes.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Yuanyuan Yang
Stony Brook University, NY·Computer Science
Develop qubit entanglement and mapping algorithms that let tens to hundreds of interconnected quantum processing units act as one virtual quantum machine with many more qubits.
Hulya Seferoglu
University of Illinois Chicago, IL·Engineering·Early-career
Characterize the cost-benefit trade-offs of coded computation, which adds redundancy to tolerate slow or failed workers, in practical edge computing systems.
Nirnimesh Ghose
University of Nebraska-Lincoln, NE·Computer Science
Develop scalable trust-establishment methods for farm wireless nodes that use chirp spread spectrum (CSS) modulation, a long-range, low-power radio technique used in agricultural IoT.
The school
Purdue University: 43 labs on PI Pages, $932M a year on research, 833 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.