Need a snack from around town? Send Spot!

[Dave Niewinski] clearly knows a thing or two about robots, judging by his YouTube channel. Usually the projects involve robot arms mounted on some sort of wheeled platform, but this time it’s the tune of some pretty famous yellow, spot-like robot legs from Boston Dynamics. The principle is simple – tell the robot what snacks you want, entirely by voice command, and it will fetch. But, we’re not talking about navigating to the refrigerator in the same room. We’re talking about trotting out the front door, down the street, and across the roads to visit the favorite restaurant. Spot will order snacks and bring them back. Completely autonomous.

Spot’s depth cameras provide localized navigation and object avoidance information
The local AI vision system manages by avoiding those pesky moving objects

There are several things going on here, all of which are quite large computational tasks. First, there’s no cloud-based voice control, like Google Voice or Alexa. The robot works on the principle of complete autonomy, which means that there is no Internet connectivity for any aspect. All voice recognition, text-to-speech and text-to-speech is done locally using the Nvidia Riva GPU-based AI Speech SDK, running on the local Nvidia Jetson AGX Oren carried by the back of the spot. A front webcam provides the audio feed for this. The voice recognition app listens for the wake phrase, then turns the collation command into text, for later replay when it arrives at its destination. Navigation is taken care of with a Microstrain RTK GNSS module, which has all the necessary robustness, such as dual antennas, and inertial fallback for regions with an irregular signal. Navigation is useless in the real world on its own, which is where spot depth cameras come in. These help avoid local obstacles, in keeping with the usual spot behavior we’ve all seen before. But what about crossing the road without tens of thousands of dollars of someone else’s gear getting run over by a passing truck? Spot’s dash cams are powered by the AI ​​Nvidia dash cam net platform which enables real-time recognition of moving obstacles such as cars, humans and anything else that might roam and get in the way. All in all, a cool project showing the future potential of AI in robotics for important tasks, like getting me a beer when I need it most, even if it’s from the corner store.

We love robots here. Robots can mow your lawn, navigate inside your home with the help of invisible QR codes, and even help you grow your food. The long-promised robot-assisted future may look more like the present.


Source link

Comments are closed.