Spot's Autonomous Adventure

2022-08-24 20:30:03
关注

The last time I asked my dog to head across town to get me a coffee she stared blankly at me, eventually deciding that whatever I said was of no interest to her. As expected, it was up to me to put my shoes on and go get the coffee myself.

Sorry, you wanted me to get what now?

Dave Niewinski, one of our favorite robotics engineers, once again set out to engineer a solution to this problem. His development platform was the Boston Dynamics Spot quadruped robot. Spot is a highly capable mobile robot, with the ability to traverse challenging terrain while carrying small payloads. With a few essential components mounted to its back, including the 3DMGQ7-GNSS/INS and NVIDIA’s Jetson Orin, Spot was ready to head across town to pick up a snack and coffee.

Spot is designed for hands-off applications like exploration and inspection, so why does it need a precise navigation solution like the 3DMGQ7-GNSS/INS? Consider this: typical GPS systems, without the benefit of RTK corrections, will get you a position accuracy of around +/- 1 meter. This might sound good enough, but when you’re trying to precisely navigate across town without walking into traffic, all of a sudden +/- 1 meter becomes far too large of an uncertainty zone to make sure you’ll get where you need to go safely. With the 3DMGQ7 and 3DMRTK navigation system, it’s possible to achieve position accuracies as low as +/- 1 centimeter!

3DMGQ7-GNSS/INS installed on Spot

Picture yourself dropping a small earring in your living room and trying to find it. Pretty tough. Now drop it in a space the size of a quarter. Much easier. Now translate this to an application like autonomous mobile robot navigation. 1 centimeter accuracy can mean the difference between getting where you need to go or falling off a sidewalk into traffic.

But what happens if Spot, or your own mobile robot, walks into a tunnel or inside a building? As Dave points out in his latest video, the precision IMU inside the 3DMGQ7 ensures uninterrupted navigation capabilities even during GPS outages.

As mobile robotics get more advanced and autonomous navigation applications become more common, the need for precision navigation solutions grows. Whether it’s guiding your robot dog across town for a coffee and fritter, or delivering packages without ever needing human input, the 3DMGQ7-GNSS/INS is the right tool for challenges that require the highest level of performance. Learn more about the 3DMGQ7 and all of our MicroStrain inertial sensors by reaching out to our talented sales and support teams. You can also view and purchase select MicroStrain sensors on Mouser.


参考译文
现货的自治冒险
上一次我让我的狗去城对面给我买杯咖啡时,它只是茫然地盯着我,最终决定我所说的与她无关。不出所料,我自己还得穿上鞋亲自去买咖啡。抱歉,你到底想让我去拿什么?我们最喜欢的机器人工程师之一戴夫·尼文斯基(Dave Niewinski)又一次着手为这个问题设计解决方案。他采用的开发平台是波士顿动力公司的四足机器人Spot。Spot是一款高度灵活的移动机器人,能够在复杂地形上移动并携带小型载荷。在其背上安装了一些关键组件,包括3DMGQ7-GNSS/INS和英伟达的Jetson Orin后,Spot已经准备好了穿越城市,去取点零食和咖啡。Spot主要设计用于无需人工干预的应用,如勘探和巡检,那么它为何还需要像3DMGQ7-GNSS/INS这样高精度的导航系统呢?请考虑以下情况:普通的GPS系统在没有RTK校正的情况下,定位精度大约为±1米。这听起来似乎已经足够好了,但当你试图在不撞上交通车辆的情况下精确穿越城市时,±1米的误差范围就变得太大,难以确保你能够安全到达目的地。而借助3DMGQ7和3DMRTK导航系统,可以实现高达±1厘米的定位精度!安装在Spot上的3DMGQ7-GNSS/INS 想象一下你在客厅里掉了枚小耳环,想要找到它可不容易。现在把耳环比作大小只有硬币般的一小块区域,就容易多了。将这个类比到自主移动机器人的导航应用中,1厘米的精度差异,可能就决定了机器人是成功到达目的地,还是掉进人行道的车流中。但如果Spot,或者你自己的移动机器人,进入隧道或建筑物内部,又会发生什么呢?正如戴夫在其最新的视频中指出的那样,3DMGQ7内部的高精度惯性测量单元(IMU)可以确保即使GPS信号丢失,也能保持连续的导航能力。随着移动机器人技术日益先进,自主导航应用越来越普及,高精度导航系统的需求也日益增长。无论你是想让你的机器人狗去城对面买咖啡和咸肉圈,还是希望它实现无需人工干预的快递配送,3DMGQ7-GNSS/INS都是应对需要最高性能挑战的合适工具。如需了解更多关于3DMGQ7的信息,以及我们所有的MicroStrain惯性传感器,欢迎联系我们的专业销售与支持团队。您还可以通过Mouser网站查看并购买部分MicroStrain传感器。 了解更多 获取报价
您觉得本篇内容如何
评分

评论

您需要登录才可以回复|注册

提交评论

提取码
复制提取码
点击跳转至百度网盘