Summary of FAST SENSE BRINGS 8X PERFORMANCE TO MOBILE ROBOTS WITH ITS REVOLUTIONARY EDGE COMPUTATIONAL AI PLATFORM
Fast Sense Studio launched the Fast Sense AI Edge Computer at Embedded world 2021 DIGITAL — a compact board-level COM Express module combining six independent AI inference engines (Intel GPU, 2x Myriad X, 3x Coral) and an Intel CPU. It runs up to six neural nets concurrently, processes multiple sensor streams in real time, includes many hardware interfaces, ships with ROS examples in Docker containers, and targets mobile robotics use cases like drones, delivery robots, inspection, logistics AGVs, and disinfection drones.
Parts used in the Fast Sense AI Edge Computer:
- COM Express single board module with Intel CPU
- Intel GPU accelerator
- 2x Myriad X accelerators
- 3x Coral accelerators
- Edge AI accelerators (set as listed above)
- Hardware interfaces for robotic sensors and actuators
- Support for multiple video camera inputs
- Software stack with ROS examples
- Docker container environment for isolated neural nets
Fast Sense Studio, a leading developer of mobile robotics solutions, today announced that its Fast Sense AI Edge Computer has been launched at the Embedded world 2021 DIGITAL.
Fast Sense AI Edge Computer is a powerful on-board computer bringing scalable Edge AI capabilities to mobile robotics.
It is a board-level circuit that integrates the power of six computers in a single module. Six independent AI inference engines (Intel GPU, 2x Myriad X, 3x Coral) operate in one compact device enabling the running of six different neural nets simultaneously without performance degradation, while receiving data from several sources in real time.
Fast Sense AI Edge Computer delivers massive artificial intelligence capabilities and has eight times more computational power compared to existing products in the market.
Thanks to six independent Edge AI engines, Fast Sense AI Edge Computer has sufficient hardware resources to run different AI algorithms and control a robot that needs to process data from different sensors like video cameras and react to changes in its environment in real-time.
The Fast Sense AI Edge Computer makes it possible to run a robot with three video cameras pointing in different directions where each video stream has to be processed with three different neural nets: object detection, semantic segmentation and depth estimation. It consists of a single board COM
Express module with Intel CPU, a set of edge AI accelerators to inference several neural nets on-board in real time and has numerous hardware interfaces to robotic sensors and actuators. The platform is shipped as a ready-to-use device with software examples of running ROS (Robot Operating System) algorithms with integrated neural nets meaningful for robotic applications, running in isolated Docker containers.
The Fast Sense AI Edge Computer can be integrated with multiple robotics platforms, e.g. drones and robots with wheels for a great number of use cases. This universal reasoning system can be applied across different classes of physical agents such as delivery robots, autonomous machines, facilities inspection, logistics robots (AGV) and disinfection drones.
- What is the Fast Sense AI Edge Computer?
It is a compact board-level COM Express module combining an Intel CPU and six independent AI inference engines to run multiple neural nets on-board in real time. - How many AI inference engines does the device include?
The device integrates six independent AI inference engines: Intel GPU, 2x Myriad X, and 3x Coral. - Can the Fast Sense AI Edge Computer run multiple neural nets simultaneously?
Yes, it can run up to six different neural nets simultaneously without performance degradation. - What kinds of sensors can it process data from?
It can process data from multiple sensors including several video cameras and other robotic sensors and actuators via numerous hardware interfaces. - Does the platform include software examples for robotics?
Yes, it is shipped with software examples of running ROS algorithms with integrated neural nets, running in isolated Docker containers. - What robotic platforms can it be integrated with?
It can be integrated with multiple robotics platforms such as drones and wheeled robots, including delivery robots, inspection machines, logistics robots (AGV), and disinfection drones. - How much performance improvement does Fast Sense claim?
The product delivers eight times more computational power compared to existing products in the market, according to Fast Sense. - Can it handle different AI tasks on separate camera streams?
Yes, it can run different neural nets like object detection, semantic segmentation, and depth estimation on separate camera streams simultaneously.
