In most instances, Vacuum robots detect collisions by switch signal. Switch is not only high-cost due to harsh working environment, but also big collision noise, and direction of the collision can only be given roughly by switch installation direction. This document presents and validates a solution for detecting collisions by 3D linear hall sensor, which is not only inexpensive, but also low noise by using flexible connections, and achieve for relatively accurate collision angle values.
All trademarks are the property of their respective owners.
Vacuum robots have had a great market growth in recent years, with an increasing number of people opting to buy a vacuum robot instead of manual sweeping. Autonomous systems, like vacuum robots, contain a lot of sensors to detect their working environment. Giving them the ability to detect things such obstacles in their path, if there is cliff on the ground, or if they have experienced a collision with something.
From high-end to low-end vacuum robots, collision detection is a basic and necessary function because the floor contains many obstacles. Even though vacuum robot has many sensors to detect its environment, sometimes collision is inevitable. If a collision occurs, the robot must not continue to move forward but turn in another direction after judging where direction of the collision.
Collison detection is necessary so that the vacuum robot can adjust and survey which direction is available. The noise from collisions can be annoying, with the source of noise not only from robot striking another object but also from the robot’s mechanical chassis.
The traditional way of detecting collisions is by using switches, mechanical switches or infrared receiver and transmitter pairs can be used as shown in Figure 2-1. To ensure that the switch signal is working reliably, the collision shell, which is part of the chassis, should compress deeply so that switch contacts may completely connect and release. As a result of the mentioned mechanical constraints, the collision noise in the event of a crash is inevitable.
In most instances, the mechanism is not simple as pictured above. These systems tend to be more complex and customized, which leads to a higher production cost. Another challenge a vacuum robot must overcome is harsh working environment. Cleaning products must operate appropriately in areas where various contaminants could be present. These contaminants include but are not limited to dust particles, oil, sewage, and even liquids such as juice or sauce. In order for the system to detect the direction of a collision, multiple switches or sensors are strategically placed to detect various directions. Using multiple sensors can lead to an increase in material and manufacturing cost.
TI offers several different types of Hall-effect sensors: latches, switches, 1-axis linear, and 3D linear sensors. Linear sensors can provide detailed magnetic data since they represent the magnetic field strength using an analog voltage or a digital numeric representation. The voltage or numeric representation changes as the field experienced by the device’s changes. Using these values, the distance can from the hall sensor to the magnet can be easily calculated. This makes linear Hall-effect sensors extremely useful for proximity sensing applications. Unlike single axis sensors, 3-axis or 3D linear sensors can use the magnetic field data from all 3 directions to calculate angle and magnitude. The three axis of sensitivity of a 3D linear Hall-effect sensor are defined as shown in Figure 3-1.
At the time of this article TI offers two types of 3D linear Hall sensors, one with an I2C interface the TMAG5273 and the other with a SPI interface the TMAG5170. Each of these 3D Hall sensors provides multiple magnetic range options that can benefit different use cases and different environments. For example, if the sensor is placed near a high-current cable, or besides running motors, a high magnetic field can be generated by these other devices. Using a different magnetic range option can yield a better SNR (signal-noise rate).
Compared with mechanical or IR switches, Hall-effect sensor are more reliable, and they are minimally affected by environment changes like temperature, humidity, dust, or water.