In recent years, are generally available cleaning robots, and robots mowing grass.
A very important and difficult issue in robotics is navigation. Each robot must know its own position. Without this information, it was impossible to reach the target, wash the entire floor or reaching to the charging station batteries. There are many evaluation method and navigation algorithms. They differ in complexity, accuracy, and of course price. Not all methods are universal and can not be used in all conditions. An important element is to build maps navigation environment. If the robot knows the own position and the position of obstacles can plan an optimal path to the goal.
 |
| My robot:) |
Please read the article.
http://www.doc.ic.ac.uk/~nd/surprise_97/journal/vol4/jmd/
Please answer the questions.
Which of navigation methods should be used in the autonomous car?
Is robot mowing the grass should have a GPS receiver?
What this map (vector, raster) should create a robot cleaning the apartment?
Which of navigation methods should be used in the autonomous car?
ReplyDeleteThis is not my field of expertise but intuitively a few methods should be combined to increase the reliability of navigation. GPS should definitely be one of the elements with the addition of the analysis of the images from cameras. Deep learning is one of the methods currently used for the interpretation of videos from cameras in the prototypes of such cars.
Is robot mowing the grass should have a GPS receiver?
GPS could help but in most situations it seems to be an overkill. Usually lawn moving robots will be used in one or a few gardens and in such situations robot will be able to build the map of terrain by itself.
What this map (vector, raster) should create a robot cleaning the apartment?
It is a technical question but my guess would be that these should be vector maps. It easier to store and analyze them so that would seem a better choice.
1. Which of navigation methods should be used in the autonomous car?
ReplyDeleteThere should be many methods used at the same time for safety. If one method fails the autonomous car should fall back on a different one. I know Google's driverless car uses GPS, very detailed maps and a rotating laser to scan the environment. Video cameras are also an important component in other solutions.
2. Is robot mowing the grass should have a GPS receiver?
I'm sure it could use GPS, but to successfully navigate through rough terrain, other methods would have to be used as well. Mowing the grass is a little more complicated than vacuuming for example (floors are mostly flat), unless we're talking about a football field.
3. What this map (vector, raster) should create a robot cleaning the apartment?
I don't know, perhaps raster would be better? Because the map doesn't have to be scalable and it's not very big, so there's no advantage when storing it as vectors.
This comment has been removed by the author.
ReplyDeleteThank you very much for your reply.
ReplyDeleteThe GPS receiver has a large margin of error.
In an automatic car GPS must co-operate with other sensors or the local station.
In my opinion, vector map advantage is its small size, scalability and multi-layered.
In the summer he wants to run my robot mower. Then check on what the map is optimal for him.
In my robot used raster maps.
Which of navigation methods should be used in the autonomous car?
ReplyDeleteA:/> AFAIK the navigation accuracy is somehow related to actual speed, magnetic factors etc.... Similar to most people I've no expertise on this matter hence my answer is perhaps the one that a regular person would say. All available :)
Is robot mowing the grass should have a GPS receiver?
A:/> I couldn't get how the grass, soil etc. are relevant to GPS receiver?
What this map (vector, raster) should create a robot cleaning the apartment?
A:/> Depends to the apartment. I saw a small robot with jumping capabilities some days ago. Imagine it climbs the stairs, you would need the vector map. BTW, raster map sounds like more processing power?
Q1. Which of navigation methods should be used in the autonomous car?
ReplyDeleteŁukasz, frankly speaking I'm not familiar with the topic. Nevertheless, I am going to do my best and give an answer. Autonomous cars move without man involvement so, it is needed to use a kind of navigation that guarantee safety for all traffic participants. In my opinion GPS with combination of cameras systems and image analysis should be enough. For me a few centimeters of diferencess in this case doesn't matter.
Unfortunately, the GPS does not always work properly. When your car is in the underground garage receiver does not work. Then the camera and laser distance sensors that can take the initiative.
DeleteQ2. Is robot mowing the grass should have a GPS receiver?
ReplyDeleteIn my opinion grass robot is destined to a local are. So , I think that the markers or tracers combined into system (at least three to set surface) should be enough to define and calibrate the local system of navigation. I think that GPS in this case is not necessary - it is only garden not a road:)
I think the open and flat space worked well to radio or infrared transmitters. In each vertex of the plot was to be the transmitter.
DeleteThe robot could easily travel around the whole area.
Which of navigation methods should be used in the autonomous car?
ReplyDeleteAutonomous cars should use both local and global navigation.
Local in terms of boundries, obstacles, pedestrians and so on.
Global in terms of route planning.
Is robot mowing the grass should have a GPS receiver?
It does need to - it's operating locally on a field. If field is constrained by the wall, laser coordination should be sufficient.
If on other hand grass mower should mow only part of the bigger unconstrained field it could use GPS to introduce boundries.
Anyhow I do not think that GPS has a use on a lawn mower.
What this map (vector, raster) should create a robot cleaning the apartment?
Raster because robot's points of references do not change, the room does not scale either.
It should be vector though if robot is an Alice from Alice in the Worderland ;)
I agree with you. Autonomous cars should integrate global and local navigation system.
DeleteIf the mower works on a small area that I agree with you.
But what if the mower runs over a wide area?
Q3. What this map (vector, raster) should create a robot cleaning the apartment?
ReplyDeleteThe answer is similar to the Q2. Local system of navigation based on markers system and should be enough for cleaning robots in apartments. I think that vector or raster maps are not necessary. But if I am wrong and should select a type of map I am for vector map because it is far better for calculations.
Unfortunately, I can not agree with you. Cleaning robot must plan your route. Must travel around the entire floor thoroughly.
DeleteThe robot should return an optimal route to the base after completing cleaning.
What if someone moves the furniture?
What if someone is not able to use the markers and the robot will be bad cleaned?
- Which of navigation methods should be used in the autonomous car?
ReplyDeleteThis is not simple question. It should use Map-Based Navigation supported by GPS system. Although other vision system should be implemented to avoid car collisions and to allow staying on the road. I am not sure but this last problem should be solved by Line Navigation. I mean detection of road. (What to do if heavy snow falls on road in winter?)
- Is robot mowing the grass should have a GPS receiver?
For robots moving on the grass (for example lawn mower) applying GPS is overkill in my opinion. There are other methods to navigate on the field presented in article.
- What this map (vector, raster) should create a robot cleaning the apartment?
Raster graphic is simple but for more accuracy I will suggest vector graphic. I remember It was used on 8-bit machines so for better CPUs it shouldn’t be too complicated.
I think the snow is not the only problem. There are many roads without marking the boundary line.
DeleteWhich of navigation methods should be used in the autonomous car?
ReplyDeleteI think that that kind of car should use many method because of security reasons. These mechanisms should be duplicated as same as it is in the aircraft in case some mechanism failed.
Some time ago I read somewhere that currently in Europe is developing a new GPS system that will be more accurate than that of your existing systems such system I think it fits perfectly for that kind cars.
Is robot mowing the grass should have a GPS receiver?
I think that is too complicated for robot which mowing the grass I think that in this particular case constructor can use less complicated and cheaper methods.
What this map (vector, raster) should create a robot cleaning the apartment?
I agree with Piotr that vector graphics is sufficient and does not require expensive equipment for their use.
Now GPS receivers cost 30$. It is not a big price for a good navigation.
DeleteI am of the same opinion. Safety is paramount. Each system must be duplicated. The black box it should be used in the autonomous car.
This comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
DeleteWhich of navigation methods should be used in the autonomous car?
ReplyDeleteThe autonomous car will have to use a combination of navigation methods and not simply rely on just one. Currently GPS offers around 2-4 meters of accuracy which is pretty bad for the autonomous car. If the car is 2 meters wide and has to follow a 3 meter wide lane there will be complications. Also GPS can't work in tunnels. Currently some cars are equipped with cameras that can track the lane lines to make sure the car is always on its lane. Apart from that such cameras or lasers that scan the terrain around could come in handy to make a local map. GPS could be used as an addition to adjust and place this detailed map on top of the underlying worldwide map. GPS is good for positioning something on a macro-scale. That's not enough for the autonomous car which needs more details on the environment surrounding it (other objects, road bumps, etc.). GPS will not warn it if there is a person passing the street.
Is robot mowing the grass should have a GPS receiver?
Unless the robot mowing the grass is an autonomous vehicle that mows the grass of an entire city or country it shouldn't. It should instead rely on more detailed local maps obtained from sensors, cameras and lasers.
What this map (vector, raster) should create a robot cleaning the apartment?
Raster map will be more efficient only in the case if the apartment and all objects placed there form some kind of squares. Then the unit will not have to use heavy amounts of processing power. In other cases though a vector map is more detailed and the robot will be more effective in cleaning the apartment but also drain more power.
I agree with you an autonomous car should use all available sources of information about the surrounding environment. GPS used by surveyors is very accurate. Surveyors use GPS receiver and the base point for the calculation of the correction - but it is stationary solution.
DeleteIf the resolution raster map is large then the shape of the rooms and objects can be any.
DeleteIf the robot has a GPS receiver then you can change the mowing area on the internet. This robot is a universal and easy to use. Imagine integrated mower with facebook :)
ReplyDeleteRobot that builds a map is easy to control via the Internet. You can select on the map where you want to clean up on the day of the week. The robot moved to another apartment can begin to clean up and build a new map without marked with special marks.
ReplyDelete1//Which of navigation methods should be used in the autonomous car?
ReplyDeleteNavigation is a huge challenge researchers and engineers must overcome when designing an autonomous vehicle. Navigation is the key to an autonomous vehicle’s success and utility. Navigation describes how an autonomous vehicle intelligently moves and interacts with its environment. Navigation can be perceived as the highest level of abstraction of motion control. There are two main focuses when designing a navigation system for an autonomous vehicle: path planning and obstacle avoidance.
For this reason I prefer use the ALM’s Navigation System in the autonomous car.
The ALM’s navigational algorithm effectively generates a path plan that covers any bounded region with any number of bounded obstacles. Several maps were input into the simulator to test various aspects of the navigation system. To date, no problems with the functionality of the algorithm have been discovered.
2//Is robot mowing the grass should have a GPS receiver?
ReplyDeleteYou should consider flipping the system around. There is no need for the robot itself to determine the location. It only needs to know what to do. This can be communicated to it from a fixed PC via a WiFi link. With such a link it doesn't matter whether the robot figures out the location or whether that is done in the fixed installation and then the result transmitted to the robot. If the robot ever looses the WiFi connection, it could simply stop. That keeps it from getting out of range and therefore not getting the information that it should turn around, meanwhile mowing down all the flower gardens in the neighborhood. I think it's also a good idea to keep the robot as simple as possible and put as much of the burden onto the fixed installation where it's easier to monitor, fix, and work with
3//What this map (vector, raster) should create a robot cleaning the apartment?
ReplyDeletewith map (vector, raster), we can't create cleaning robot because the navigation course keeps his methodical mode using a laser Infrared which scans its environment in real time and creates a virtual map.