Three founders with 10+ year publication records, eight researchers, one goal: robots that can handle the physical world.
Grasping · perception · control · biomedical — one bench
01 — What we work on
All research →P-01
The program studies grippers and artificial fingers driven by shape memory alloy wires: their mechanism design, dynamic models and fuzzy PID control of position and grip force despite actuator hysteresis. Alongside this, computational mechanics work models layered and graded composite structures with layerwise finite elements, and non-Newtonian fluid flows with smoothed particle hydrodynamics.
People — Vahabi · Jedari Salami · Khodayari
P-02
The program studies machine perception from physiological signals, speech and medical images: detecting sleep arousals in EEG, predicting seizures from heart-rate variability, recognising emotion from ECG and speech, and segmenting coronary arteries in X-ray angiograms. Its questions are which representations expose subtle patterns, and which learning methods detect them reliably.
People — Farokhi
P-03
Control of systems whose dynamics are only partly known: wheeled robots, quadcopter formations, magnetic levitation, electric motors and grippers driven by shape memory alloy wires. The program studies adaptive, sliding-mode and fuzzy controllers, online neural-network identification, data-driven models of human car-following, and multi-objective routing, scheduling and supply-network design under uncertain costs and demand.
People — Khodayari · Barzamini · Kazemipoor
P-04
The program uses finite-element and fluid-structure interaction models to study how a knee ligament and prosthesis, a heart valve, the airway and the head respond to load and flow. It also studies machine-learning methods that detect sleep arousals, predict seizures and support diagnosis from EEG, ECG and medical images.
People — Kashaninia · Farokhi · Shojaei
02 — How we work
A hardware result is announced only when it works on the bench, with the video of it working posted next to the claim.
Every publication listed carries a DOI and the affiliation it was published under. Faculty records link to Google Scholar and ORCID.
Where evidence does not exist yet, the page leaves the claim out rather than writing a plausible sentence. That is policy, not accident.
DOI-linked. Every entry was published under its authors’ university affiliation, printed on the row; none was published under the institute’s name.
04 — For engineering teams
Bin picking that jams. Parts that deform when gripped. A cell that needs re-fixturing every batch. We take these as scoped engagements: a feasibility study with a written answer — what works, what does not, what it would take — then, if the answer is yes, a prototype with acceptance criteria. If physics says no, you get the no in writing, with reasons.
Scoping call. Problem, constraints, success criteria. No cost.
Feasibility study. Fixed scope, written report.
Prototype. Statement of work — deliverables, dates, acceptance criteria.
05 — Contact
Contact the person you would like to speak with about research or collaboration.
Choose a recipient. Your email app opens a draft for you to review and send.
Tehran, Iran
Meetings by appointment. Please email your contact to confirm the time and location.