Lynx M20S: Chinese robot dog faces rivers, ice and -30ºC cold

Since Boston Dynamics launched BigDog - one of the first robot dogs aimed at real applications - in the mid-2000s, a lot has changed in this area. Current models, with articulated wheels and legs, have evolved so much that they look more like autonomous vehicles than these experimental quadrupeds.

Recently presented by the Chinese DEEP Robotics, the Lynx M20S model is defined as a "wheeled-legged robot" (robot with wheels and legs) designed to face extreme cold, hostile terrain and missions where humans have difficulty performing.

During tests, the robot crosses frozen surfaces and also crosses flooded areas up to 80 centimeters deep. Operation is maintained even at temperatures down to -30 °C. At other times, it can face slippery slopes of 45 degrees without losing stability.

Held at altitudes close to 5,177 meters - a range in which rarefied air and extreme cold make human action difficult -, the presentation sought to reproduce the environment that defines the practical use of the robotic system, according to DEEP Robotics.

The Hangzhou company also recently presented the DR02, a humanoid robot aimed at industrial and firefighting operations. The consecutive launches indicate the company's strategy to expand its presence in field robotics.

What are the technical specifications of the Lynx M20S?

DEEP Robotics Lynx M20S | The Ultimate All-Terrain Champion.

Part of the performance presented comes from the hybrid architecture. The Lynx M20S can alternate between wheels - faster on flat terrain - and articulated legs, used to overcome obstacles and irregularities. This results in a robot that is much more versatile than its leg-only counterparts.

Weighing 35 kilos, the Lynx M20S was designed to operate in narrow spaces, just 50 centimeters wide. The robot climbs 25-centimeter steps, overcomes obstacles of up to 80 centimeters and faces slopes of 45 degrees. The robot carries up to 15 kilograms of payload and supports a maximum load of 50 kilograms.

Its autonomy reaches three hours or 15 kilometers without a charge, a time that reduces to 2.5 hours or 12 kilometers when it has a maximum load on board. To orient itself, the Lynx M20S uses two 96-line LiDAR sensors, that is, 96 laser beams combined to map the relief in high vertical resolution.

Complete the set are wide-angle cameras that provide a 360 by 90 degree field of view, which allows you to avoid obstacles in any surrounding direction. The model comes with IP66 certification, which ensures total protection against dust and resistance to strong jets of water.

What is the Lynx M20S for and who are its main competitors?

The Lynx M20S was designed to perform inspection of electrical networks, emergency response, logistics and scientific exploration - DEEP Robotics/Disclosure
The main "mission" of the Lynx M20S will be to perform inspection of electrical networks, emergency response, logistics and scientific exploration. Among the scenarios envisaged for action, the manufacturer cites mountain trails, flooded areas, tunnels, pipelines and regions affected by disasters.

The launch takes place at a time of effervescence in Chinese industrial robotics, in which companies like Unitree have been gaining ground. Not to mention Western heavyweights such as Boston Dynamics and Agility Robotics, which compete in the same niche of robots for difficult terrain and risky operations.

Confidently, DEEP Robotics ensures that the combination of wheels and legs extends autonomous operations into environments currently inaccessible or dangerous for human workers. In addition, the proposal reduces people's exposure to extreme cold, high altitude and unstable terrain.

With no estimated arrival date in Brazil nor a published price, the Lynx M20S joins the list of models that mark the advancement of field robotics. The current demonstration reinforces a question that already accompanies the sector: what will be, in practice, the limit of human replacement in risky missions?

From camera to three-dimensional map: how robots learn to see

Source: CNN