Huawei and Deep Robotics:
Are the perfect team to enable the JY Mini AI Robotic Dog, to undertake Patrol and Inspections in industrial environments.
On 25th September 2020，during the keynote speech on stage at the Huawei Connect Live 2020 Conference, hundreds of thousands of eyes (in house and online) were staring at the cute robotic dog.
JY Mini looked calm and happy, but actually, it had a tough mission ahead!
The Mission
In a factory with many pipelines, a pipe leak somewhere, has caused a local fire and all kinds of alarms have not yet started to call out. As an intelligent robotic dog on patrol, can JY Mini respond quickly and resolve the crisis in time?
Huawei ModelArts has co-operated with Hangzhou Deep Robotics Company Ltd, to train quadruped robots with its core Huawei cloud EI-backbone, to create the technology for JY Mini to carry out inspections in an industrial setting.
JY Mini robot dog has the ability to identify, judge and deal with crises in real time, which is a perfect complement to its existing "mobile omnipresent" movement ability.
In the case of a fully autonomous operation, JY Mini can quickly detect a flame, understand that the fire is harmful to the factory environment, relearn and plan to generate a new path to bypass the flame and turn off the control valve to help put out the source of the fire, as illustrated in the pictures below.
Professor Zhu from Deep Robotics and Zhu Jiancheng from Huawei (pictured above) introduced further details about the co-operation between the two companies. The use of ModelArts and Atlas 200DK has given the Jueying "JY" Series of robot dogs AI capabilities, to be able to undertake intelligent inspection of factory settings. The JY Mini can perceive the on-site environment in real time and strengthen the learning dynamic decision-making through the interactive analysis of a knowledge graph. JY Mini also has the capacity for complex path planning and action and through the use of Huawei Cloud El–backbone technology, so JY Mini can ultimately protect the safety of the factory.
Zhu Jiancheng said “JY Mini achieved a qualitative leap in intelligence through ModelArts, by building the abilites of "perception + cognition + decision making" for the JY Mini robotic dog.
Professor Zhu believes that this "simulation exercise for factory inspection and fire fighting" highlights the fact that mechanical engineering, cybernetics and computers are integrating with AI technology. The intelligence of the JY Mini quadruped robot is breaking through perceptual interaction and moving towards the stage of cognitive decision-making. With the continuous integration of AI technology, the Jueying (JY) “shadow”, will show more amazing abilites in the future!
