Robotics and Artificial intelligence showcase was held in Manchester, 12th of March 2019.
Radio links
The Oil & gas technology centre had a really nice point cloud of oil rig, down to the minute detail (video below).
The centre provides an umbrella for a number of smaller companies, one of which is the drone operator Air control entech. Air control entech operates large drones,, having operations and drone pilots.
Clogworks manufactures drones (with distinct aesthetics), and CEO provided very interesting insight in radio link dependency. In particular, the main competition is from DJI with Lightbridge 2 radio link (5km range). Clogworks is happy with their new radio link 20km, 25Mbps (in co-operation with Sky-drone). The next step for Clogworks is further increasing the range and size of their drones (100kg payloads, e.g.).
SMC corporation from Tokyo has an industrially used radio link (one master, 6slaves, 250kbps, frequency hopping – for security reasons, 5ms delay) EX600-W. The links is used for robotic arms where due to the movement the data link cable gets broken often, or where the robotic arm needs a full 360degrees movement at the pedestal. SMC is the only vendor of wireless solutions for robotics in the UK.
Robotics and AI in nuclear (RAIN) is of potential interest for small cells community and they will have:
- a demo day on 26th March
- assurance on 12th April (NB: to provide more info here)
- game changers on 19th September (NB: to provide more info here)
Robotic arms
University of Liverpool has active co-operation with Kuka robots, and has won some awards from Kuka challenges.
ABB has reserves with wireless communication from the safety – risk point of view. ABB does not, as of 2019., manufacture moving robotic platforms.
Automata exhibited their Eva robotic arm for less than £5,000.
Computer vision
Manufacturer of European Mars rover, SCISYS are moving away from nVidia (clearly for radiation hardening reasons).
London based, Sensing feeling has a low power solution (Jetson TX2 –> Xavier, and Intel Atom + Movidius VPU platform – I assume Myriad X) for face and feelings recognition.
(There is a lingering but not confirmed assumption that FPGA solutions have less delay/power comparing with GPUs.)