The world is in the midst of a robotics revolution, and the humble vacuum cleaner is leading the way. It is no surprise then that the number of robot vacuum cleaners available to consumers has increased almost exponentially as new entrants have entered the market and competition heats up. This is great for pushing robotics technology forward. Indeed, the number of features on these obedient machines and their level of intelligence continues to increase. However, the rapid pace of change can also make for a confusing picture. To help you decide which robot vacuum cleaner is best for you, we have looked at the range of robot cleaner offerings in detail. The focus here is on the current state of the technology and determining the key features to pay attention to when purchasing one.
-- information not provided by the manufacturer
What to look for when buying a robot vacuum cleaner
Probably the most important aspect of any vacuum cleaner is its ability to pick up dust and dirt. Its suction power or pressure is often measured in Pascals (Pa). As one might expect, the higher the number of Pascals, the greater the suction. However, increased suction also means that a robot vacuum's batteries will drain faster. Higher suction is usually needed on carpet, and the thicker the carpet or rug, the greater the power required. Consequently, most robot vacuum cleaners provide a good clean on low to medium pile carpets but struggle with thicker floor coverings. Only the higher suction machines are able to handle tougher types of flooring.
Importantly, many robot vacuum cleaners automatically adjust their suction power when moving between hard floors and floor coverings. These machines will increase their suction power when moving onto carpets or rugs, while reducing it again when re-entering areas of hard flooring. This feature extends battery life without compromising on cleaning quality.
After suction, navigation is the next most important feature to pay attention to for robot vacuum cleaners. There are several different methods that robotic vacuum engineers employ to move their robots around in a given space. In increasing levels of sophistication, these are listed below:
As the name suggests, robot vacuums in this category travel an entirely random route. Unlike more sophisticated machines, they do not create a map of their workspace. Therefore, they never really know where they are within their surroundings. They usually move in a straight line until they come up against a wall or obstacle before changing direction.
The downside of random navigating robot vacuums is that they tend to be more susceptible to becoming stuck. The classic case is when they clean under dining tables where the high concentration of chair and table legs can trap a visually-challenged machine! On the other hand, vacuum cleaners in this category are the least sophisticated and consequently the least expensive. As a result, they are ideal for consumers on a budget.
Robot vacuum cleaners of this type use gyroscopes to provide a robot with a sense of relative direction. These are often combined with optical flow sensors which sense the distance the machine has moved in a single direction. Together, these sensors usually allow a robot vacuum to generate a crude ‘map’ of its surroundings. Mapping technology lets the user know where the robot has cleaned and the areas where it may have missed. However, this type of mapping does not usually allow the user to define zones or rooms to work in nor to avoid sensitive areas. Consequently, gyroscopic navigation in robot vacuums is not very popular today and is unlikely to be used for robot cleaners of the future.
In contrast to gyroscopes, LiDAR-based technologies have become the dominant form of navigation that advanced robot cleaners are using today. LiDAR uses a laser to detect and range objects in the robot’s surroundings. It does this by spinning the laser around the room, capturing the light reflections as they bounce off obstructions in the space. Using this reflected light, the robot is able to build up a picture of the room, creating a more sophisticated map by which to navigate. With LiDAR-based mapping, a robot not only knows where it has cleaned but can also be directed to clean certain zones or rooms. In addition, no-go areas can be defined to prevent it from disturbing sensitive locations and avoiding areas where it may get into trouble. LiDAR is now the most favoured navigation technology in top-tier robot vacuums since it tends to be more accurate than other types of mapping technologies.
Another type of navigation technology that some of today's robot vacuum cleaners use is vSLAM. Instead of gyroscopes or lasers, vSLAM-based robots uses cameras to ‘see’ their surroundings and to identify points within it. The robot uses these navigation points to locate its position within a space. In addition, it is also able to generate a map of the workspace. As with other types of mapping, a user can use these maps to know where the robot has cleaned. Additionally, the maps can also be used to direct a robot to work in certain areas or rooms while avoiding others.
Currently, vSLAM occupies a minor position compared to LiDAR, when it comes to the preferred technology for robot vacuum cleaner navigation. This is mainly due to LiDAR's increased accuracy and better mapping capabilities. However, the use of cameras on robots means that they can assimilate a greater amount of data from their environment than LiDAR. This may not seem as important in today's robots but as AI advances, robots will increasingly make use of cameras to understand their environments. Indeed, some of today's advanced machines already have cameras on their front face. These are used for enhanced AI-based obstacle avoidance (more on that below). As a result, vSLAM-based navigation may well become the technology that robot vacuum cleaners will use in the future.
Boundaries and Zones
Robot vacuum cleaners often have to be directed away from sensitive locations where they may inadvertently damage something sensitive or get into trouble in some way. Consequently, virtual walls are often needed to delineate these areas. In addition, directing a robot cleaner to focus on cleaning certain rooms or zones is a feature that most of us find useful. Therefore, most of today's mapping robots also often feature room or zone labelling. These virtual boundaries and zone designations can come in various forms as described below.
The least sophisticated virtual walls come in the form of physical magnetic strips. These are usually placed on the floor, often under rugs or carpeting, to stop a robot cleaner from crossing into a particular area. For the magnetic strips to work, the robot vacuum cleaner also has to have the sensors to detect them. The major disadvantage to using magnetic strips is the inconvenience of placing them, especially if not permanently installed. In addition, if not concealed, their rather unsightly appearance lying around on the floor is obviously unappealing. Finally, the number and size of the strips that come included with a machine are limited. So if lots of areas or rooms need cordoning off, more will need to be purchased which will mean greater expense. Fortunately, this older form of boundary technology is rapidly going extinct as the machines of today become more sophisticated in their navigation.
Another form of essentially-extinct virtual wall comes as a stand-alone device. These boxes emit infrared light to mark out a space where a robot should not enter. They can often be set to emit different light patterns, for example, a straight line or a circular path. The straight line configuration creates a straight virtual barrier to stop the robot from crossing into an area, useful for blocking doorways or narrow corridors. Meanwhile, the circular pattern is useful for encompassing a sensitive area protecting objects within a specific spot. One obvious disadvantage of these stand-alone boxes is the requirement for batteries to function. In addition, their physical nature limits the number of areas that can be cordoned off to the number of devices available.
No-Go lines & Zone cleaning
Today, the predominant type of virtual wall is one that can be set directly in a robot-generated map. Of course, the robot has to use advanced mapping technology in the first place to provide this functionality. A user can indicate on the map which areas the robot should avoid as well as direct the machine to clean certain zones or rooms when required. However, the main disadvantage of this type of virtual fencing is that it is usually only found on more expensive robot vacuum cleaners with advanced navigation abilities. Random and gyroscopic-navigating robots usually do not possess this capability.
In addition to navigating, robot vacuum cleaners also have to recognise and react appropriately to obstacles in their path. To this end, various obstacle-avoidance technologies have been developed over the years with the level of sophistication rising with every generation of machine. These are examined below.
Today, all robot vacuum cleaners, including the most sophisticated ones, have a bumper on their front face. The bumper lets the robot know when it has bumped into an obstacle. However, the bumper's level of importance in a robot's obstacle avoidance capabilities is different for machines of different sophistication levels. In the case of randomly-roaming robots that do not employ any other means of navigating, the bumper plays a central role in determining the ultimate path of the robot cleaner. However, for robots with additional obstacle avoidance technologies, it tends to act as a fall-back obstacle sensor that is activated only when other technologies fail to recognise an obstruction.
The next level up in obstacle avoidance technology are infrared detectors. These are designed to detect infrared light emitted from the robot which has reflected off the surface of an obstacle in the robot's path. In this way, a robot can avoid touching an obstruction completely and avoid any risk of trauma to either it or the obstacle. However, infrared light is not reflected sufficiently from every surface, with black surfaces tending to absorb the light rather than reflect it. As a result, infrared detectors often work hand-in-hand with the front bumper to avoid the full range of obstacles a robot may find in its environment.
More sophisticated robot cleaners emit structured light - think light pattern - which reflects off the surfaces of obstacles in a robot's path. This is not dissimilar to how infrared detectors (just discussed) work. Indeed, the light in structured light technologies is itself usually infrared light. However, with structured light, the robot is able to analyse distortions in the projected pattern that allows it to perceive shape information of an obstacle in its path. This improves obstacle recognition and avoidance over simpler infrared sensors.
Finally, we come to the most advanced form of obstacle avoidance, robot vision. Using one or two cameras, a vision-based robot vacuum cleaner will take a continuous series of pictures of its way ahead. Then using artificial intelligence algorithms, it will analyse them to determine what is or is not a potential obstacle. In most cases, the robot is able to actually identify what the object is, matching it to a small database of known object categories. This lets the robot react more appropriately to specific obstructions. Furthermore, the AI algorithm can often also update itself about the objects in its particular environment further improving its obstacle avoidance mechanisms.
Many of today's robot vacuum cleaners also have the capability of mopping the floor in addition to vacuuming it. Robot mopping usually involves the application of a small amount of cleaning fluid to the floor just prior to a sponge pad wiping it down as the robot moves forward. Importantly, most mopping robots can detect carpeted flooring or the presence of a rug which, for obvious reasons, should not be subjected to the cleaning fluid or the mopping pad. In such situations, the robot will either turn away or will stop applying fluid and lift up its mop to prevent it from soiling the floor covering as it passes over it.
Robot vacuum cleaners usually have more than one way to activate and control them. At the lowest end of the technology spectrum are physical buttons on the machine itself. In day-to-day use, the robot vacuum cleaner usually runs on a schedule or in response to app-based commands from a mobile device. However, sometimes one wants to quickly activate the machine without having to go through the hassle of opening an app on a digital platform. Consequently, many machines incorporate physical buttons on the machine itself which can be used to start and stop the machine manually.
The next level up when it comes to controlling the robot vacuum is the dedicated remote control, not unlike the traditional TV remote. These come with some robots and can be useful if the more high-tech internet-based control methods (discussed next) are unavailable. With a remote control, one is usually able to activate the robot as well as set its cleaning mode. In addition, most remotes also provide a ‘manual driving’ facility so that the robot vacuum cleaner can be directly steered if needed. Importantly, most robot vacuum cleaner remote controls only work when in direct line-of-sight of the robot.
The mobile device-based system is the most advanced of the robot cleaner control systems. It works by connecting the robot cleaner to Wi-Fi which then mediates its connection to the robot manufacturer's app. Robot cleaner apps usually sport additional functionality such as setting cleaning schedules and receiving notifications if the robot gets into trouble. The use of a Wi-Fi-connected device to interact with the robot cleaner also usually allows control of the robot from locations away from home. For mapping robots, app control is critical to the functioning of the robot cleaner, allowing interaction with the cleaning maps of one's home. Apps also usually provide the same functionality as a physical remote control allowing users to activate the machine, as well as manually steer it from the mobile device. However, it should be noted that due to the extended data transmission route of app-to-cloud-to-robot, manual steering via an app is usually not as responsive as the more direct path of a physical remote, so 'driving' tends to be a bit more difficult!
Finally, most machines that can connect to Wi-Fi are also able to integrate with a smart speaker such as an Amazon Echo or Google Home device. This allows the user to activate the robot cleaner or send it back to its charging dock simply by instructing Alexa or Google Home. In lower-end robot vacuum cleaners, this will simply consist of telling the smart speaker to start and stop the robot. In higher-level machines, users are also able to specify the room or location in which to do the cleaning.
Today, nearly all robotic vacuum cleaners come with lithium-ion-based batteries similar to those found in laptop computers. However, much more energy is demanded from a robot cleaner’s battery than that of a laptop. Running the wheel and vacuum motors in addition to the internal electronics can prove quite taxing on a robot's battery. This means that robot vacuum cleaners can usually only run continuously for about 1-3 hours on a full charge (unlike laptops some of which can go for 12 hours or more). Higher vacuum suction pressures can also drain the battery faster so one usually aims for running a machine on its lowest power setting required to provide a good clean. In addition, 'carpet boost', discussed in an earlier section, where the robot only increases suction power when it encounters carpets or rugs also helps to conserve battery power.
Recharge and Resume
Practically all robotic vacuum cleaners will try to return to their docking stations to recharge when their battery levels fall below a certain level. Random navigation machines will systematically search the area for their docks until they find it. This is inefficient and can sometimes result in them failing to find their charging home. In contrast, mapping robots will often make use of their mapping functionality to more efficiently locate their charging station. Furthermore, for the simpler, less expensive machines, a low battery charge will often signal the end of their cleaning cycle. However, more intelligent mapping robots, cleaning a prescribed area, may return to the dock just to recharge before going back out again to continue cleaning from where they left off.
Due to the size and shape of these robotic vacuum cleaners, most will invariably have limited onboard dustbin space, especially when compared to conventional vacuum cleaners. So, when choosing a robot vacuum cleaner, the bigger the bin volume, the better. However, it can be assumed that most robot cleaners will need to have their bins emptied after almost every run for optimal cleaning performance.
More recently, the latest cleaning machines are now able to automatically empty their waste receptacles while at their docking stations. To facilitate this, the docking station incorporates a separate larger stationary bin. The dust and debris is then automatically transferred to this bin from the mobile cleaner each time it docks. Eventually, the docking station bin too will need emptying but the difference here is that its volume is more comparable to that of a conventional vacuum cleaner. As a result, it only needs emptying over the course of days and weeks rather than after each robot run.
One particularly important physical dimension of any robot vacuum cleaner to keep in mind is its height. One of the advantages of having this type of vacuum cleaner is that it can work independently, going under furniture and other home fittings without having to move them. Most robot vacuum cleaners are relatively flat machines but it is always a good idea to ensure that any potential new robot purchase will have sufficient clearance under all of your home furnishings. Importantly, a lot of furnishings may have sagged (or sag when sat on), sofas being the classic example. So ensuring there is sufficient clearance throughout the underside of a furnishing is also essential.
When considering the purchase of a robot vacuum cleaner, several key factors should be taken into account. Firstly, the suction power, measured in Pascals, determines the cleaner's ability to pick up dust and dirt, with higher suction generally required for thicker carpets. Carpet boost technology adjusts suction automatically, optimizing battery life without compromising cleaning quality. Navigation is crucial, with LiDAR emerging as the favoured technology for accurate mapping and efficient cleaning. Boundary delineation, obstacle avoidance mechanisms, mopping capabilities, and control options, including app and voice control, are also important considerations. Battery life, dustbin volume, self-emptying functionality, and the cleaner's height are additional factors to weigh, ensuring the chosen model fits both cleaning needs and home layout effectively. Overall, selecting a robot vacuum cleaner that excels in these areas promises convenience and efficiency in household cleaning routines.