R&D at GPS runs on more fronts than an update can capture, from reactor science, enzymes, and byproduct recovery, to AI assisting all of it. Here’s a look at five major themes we are working on:
The Biogas Expedition
Dr. Pratibha Baral &Dr. Subodh Kumar with Rakesh, Shreyas, Siva, Rupesh
The team tests diverse feedstocks and enzymes across 24+ digesters at plant scale. This collaborative, failure-positive effort aims to unlock greater biogas potential through rigorous, scalable experimentation.
A test only fails when equipment or procedure stops providing reliable data. Everything else is learning.
Dr. Pratibha Baral & Dr. Subodh Kumar
Building our own enzymes
Dr. Nitin Kumar & Dr. Goutam Pawaskar with Amrita, Usha
Commercial enzymes often fail on local feedstocks. To lower costs and improve performance, we build custom enzymes by developing and testing resilient fungal variants. Each candidate is tested through rigorous lab-to-plant scaling gates, optimising conditions to boost methane output while respecting biological limits.
Give the strain the right conditions, and it rewards you. The better we support its growth, the more enzyme it delivers.
Dr. Nitin Kumar & Dr. Goutam Pawaskar
Beyond gas: Silica & lignin
Dr. Prathamesh Wadekar
Why stop at biogas and FOM? This process pulls silica and lignin, in high demand in rubber, construction, batteries and cosmetics, from inorganics & unused carbon, adding value without competing with gas. Now at 100L, scaling to 20 TPD+ plants.
Silica is the primary target; lignin is a promising runner-up that still needs work to scale.
Dr. Prathamesh Wadekar
Ranking the mutations
Bharath Rao K N
A platform maps reaction pathways, finds the bottleneck enzyme and ranks candidate mutations, predicted, not hand-designed, and sends them straight to the enzyme team. The limit isn’t compute; only a lab test confirms a mutation.
Its strength today is AI-assisted shortlisting, with a scientist making the final call.
Bharath Rao K N
One source of truth
Ayush Kumar
With 25 digesters in parallel, data must be consistent, traceable and validated. pH, TS, OLR are checked before any comparison is trusted. A shared repository surfaces cross-team patterns; manual checks catch what automation can’t.
No parameter exists in isolation. The insight is in the relationships between feedstock, OLR, pH and TS.
