Navigation, localization and guidance for AGVs and AMRs
Javier Miguélez
Top Robotics Voice ? Consultant for AGVs & AMRs @ Moving Robots | We help you with your logistics automation | 25.000+ followers
Lately I am seeing that there is a lot of confusion in mobile robotics customers and providers about three concepts: navigation, localization, and guidance. Let’s shed some light:
Navigation: how to move from point A (with coordinates x1, y1, z1, w1) to point B (with coordinates x2, y2, z2, w2).
Localization: where the mobile robot is. In other words: coordinates. That is: x, y, z and w (orientation of the robot). The most common localization technologies for mobile robot localization are:
?Guidance: a mobile robot following a line. Main types of guidance technologies:
?In addition, acronyms create more confusion:
Let’s see if I can help by answering some FAQs:
Yes, there is. A mobile robot can localize itself by using magnetic tape or simple magnets. Quite uncommon these days, but it exists.
领英推荐
Because guided mobile robots are usually less expensive. In addition, in many cases guidance is simpler than SLAM localization + navigation.
In most of the cases, yes. For example, with magnetic or optic guidance it is common to use RFid tags so that the mobile robot knows where to stop.
It means navigation that uses coordinates coming from SLAM localization.
Because you need to install the guiding reference and in most of the cases that is expensive and not flexible enough.
Yes, it is. Actually, most of the SLAM localized mobile robots use odometry as well.
?
If you want to add some more questions to the list, let me know!
simple is better | mobile robots | Executive Director @SAFELOG
2 年Javier Miguélez Thank you for sharing ???? knowledge is the key ??
Agile & Metal Additive Manufacturing
2 年Maybe this online course might help the italian audience: https://www.skills4i.com/it/prodotto/robotica-autonoma/
CEO & Founder ICRA - Impulsor de la robótica en Colombia y el mundo.
2 年Great!
Business Director & Co-founder at Beta Robots. PhD roboticist.
2 年Thanks for the post! Some comments just to boost the Friday chat: 1) When referring to localization that uses a map, we can call it "map-based localization", instead of SLAM, since, in general, there is no longer need of rebuilding the map, so it is not necessary to compute the mapping problem. 2) As you point out, fusion of localization techniques is always a good approach to obtain more robust solutions 3) Regarding AGV, I've also heard around all combinations of: - Automated, Automatic, Autonomous - Guided, Ground, - Vehicle Best!
安全高效的搬运解决方案
2 年Great post Javier Miguélez. Despite the drawbacks of 'guidance' options I do think they offer a cost-effective way to step into automation.. particularly line followers