These are some of the projects that AESEMI members are working on.
Robotic Processing Unit
Acceleration Robotics‘ Robotic Processing Unit is a robot-specific computational processing unit that uses hardware acceleration and efficiently maps robotic computations to its CPUs, FPGAs and GPUs for best performance. In particular, it specialises in optimising the execution of robotic computational graphs related to the Robot Operating System (ROS 2).
RobotPerf is a benchmarking suite for measuring the computational performance of robotic computations, devised by Acceleration Robotics. RobotPerf provides a set of open benchmark benchmarking tests that can be used to evaluate the computational performance of different robotic solutions fairly and with the Robot Operating System (ROS 2) as a common base, so that robotic architects can make informed decisions about the hardware and software components of their robotic systems.
This Acceleration Robotics project seeks to design and develop an open software and hardware robotic processing unit in the form of a microcontroller (MCU) with RISC-V technology that offers lower latency and additional real-time capabilities in robot operating system (ROS 2) interactions. The ultimate goal of the project is to design an MCU containing a native ROS 2 hardware implementation.
Blue Box Porbeagle Project
Project for the EU’s Space Programmes Agency, led by ArXiTEC.
Democratized Cryo Electron Microscopy Image System (DeCEMIS)
The DeCEMIS consortium, coordinated by IMASENIC, is working on the world’s first low-energy detection system optimised for electron cryomicroscopy (CryoEM).
iC-Malaga is currently working on the manufacture of passive ionising radiation sensor microchips for medical, industrial and defence applications, which it markets under the Sealicon brand.
IC KDPOF KD1053
The KDPOF KD1053 transceiver IC addresses three target markets, with individual products tailored to each specific need: automotive, home and small office plus home office (SOHO) and industrial networks.
KDPOF has developed the new integrated KD9351 fibre optic transceiver (FOT), which further reduces the cost of in-vehicle optical networks to 1 Gb/s by incorporating transmit and receive optoelectronics in a single component.
CLUE has developed a series of generalist aviation computers for aerospace and mission systems. It is based on a modular architecture with pre-certified aerospace standards.
Developed by CLUE, this is a new custom-format avionics and flight computer.
At Peta Optik they are experts in the execution of FPGA-based deep learning using INTEL solutions. Applications are in domains where speed is very important and can range from new modulation/demodulation schemes to Artificial Intelligence applied to video analysis and other vision applications.