Skip to content Dr Martin Peniak Robotics · Cameras · Coordinates · Place

Dr Martin Peniak

Pioneering GPU computing for cognitive robotics, then applied vision and spatial systems.

Dates, roles, and technical areas kept concise. The fuller narrative lives on Story, and the documentary record lives on Sources and Talks.

Profile

Early GPU robotics, deployed vision, and spatial systems.

Engineer and researcher whose University of Plymouth PhD record is described as the first investigation of GPU computing’s impact on cognitive robotics. The wider path spans iCub humanoid learning, CUDA and Aquila, evolutionary rover research with ESA, a 2012 NVIDIA research internship, Cortexica machine vision and edge AI, and continuing research and innovation through AnyVision, Oosto, and Metropolis.

Primary records

Six quick anchors.

Experience

Timeline.

Senior Innovation Engineer, A-Team – AnyVision, Oosto, and Metropolis

The same research-and-innovation role continued as AnyVision became Oosto and Oosto was acquired by Metropolis. The public record stays at a high level: applied computer vision, edge systems, spatial intelligence, synthetic vision, and technical prototyping.

Head of Innovation – Cortexica Vision Systems

Research-to-working-systems innovation across machine vision, edge AI, mobile inference, AR/VR prototypes, and applied computer vision, including the public Prism and OVERPaSS project records.

Senior Parallel Computing Software Engineer – Cortexica Vision Systems

GPU/OpenCL/CUDA optimization, high-performance image-processing pipelines, deep learning experiments, cloud systems, and edge deployment work.

Associate Lecturer of GPU Programming – University of Plymouth

CUDA and parallel-programming teaching, GPU computing lab work, and research infrastructure for cognitive robotics, including the early NVIDIA CUDA Teaching Center contribution at Plymouth.

NVIDIA research internship – Silicon Valley

Summer research internship at NVIDIA in 2012, following the 2011 Santa Clara HQ presentation of GPU cognitive-robotics applications. The wider 2011-2014 public record includes the NVIDIA CUDA Spotlight, SC11 keynote reference, GTC 2012 poster, and GTC 2014 presentation.

Cognitive robotics researcher – University of Plymouth

PhD and postdoctoral research across the iTALK and POETICON++ projects, iCub action-language learning, recurrent neural networks, Aquila, and GPU-accelerated robotics.

Mars rover research – University of Plymouth and ESA ACT

The final-year rover simulator led into the formal Evolution in Robotic Islands study. Plymouth lists Martin Peniak with Angelo Cangelosi and Davide Marocco as project leads; ESA records him among the study participants.

2012 NVIDIA intern badge for Martin Peniak

NVIDIA

An object from the CUDA robotics years.

The original 2012 NVIDIA intern badge sits beside the public sources from that period: the invited Santa Clara HQ presentation, NVIDIA CUDA spotlight, GTC poster on GPU-accelerated action acquisition, GTC 2014 presentation, and SC11 keynote reference. Together they connect the Plymouth iCub research to the early GPU-computing period.

Education

Academic foundation.

2014

PhD – GPU Computing for Cognitive Robotics

University of Plymouth. Cognitive robotics, iCub humanoid learning, action-language acquisition, GPU/CUDA acceleration, complex neural networks, and computer vision.

2006-2007

Life in the Universe

University of Manchester distance-learning astronomy course completed during the Plymouth work-study period.

1999-2003

Stredná priemyselná škola strojnícka, Zvolen

Mechanical-engineering secondary school in Zvolen, Slovakia. Its former-student records are now held by Stredná odborná škola technická in Zvolen.

Current school and records centre
Martin Peniak with the iCub humanoid robot during robotics research

Technical areas

Selected technical areas.

  • iCub action-language learning and multiple time-scales recurrent neural networks
  • GPU acceleration and multi-GPU back-propagation through time for cognitive robotics
  • Aquila and modular CPU/GPU tooling
  • Edge computer vision under physical constraints
  • Multi-camera spatial calibration and place-aware records
  • C++, CUDA, OpenCL, OpenCV, Python, and local tools

Site map

The CV connects back to the work.

The CV stays compact; the sources, talks, and work pages carry the deeper material.