Posted on

Position sensing is all about seeing black

Light-based proximity sensors for use in robots and 3D printers must not be fooled by transparent or black surfaces, says Gabriele Fulco

Figure 1: Reflective microsensor

Figure 1: Reflective microsensor

A surprising number of systems depend on the ability to accurately locate and identify physical objects. Robots need not only to detect obstacles, but also to identify the type of floor they are on, and to be aware of steps before they fall off them.

3D printers and IP cameras each present a whole new set of challenges. Vending machines need to handle a growing number of different types of packaging, including transparent glass and plastics.

The role of sensors

Light-based proximity sensors, alternatively known as photo‑microsensors (Figure 1), slotted switches, opto-switches or optical switches and photo interrupters, are generally used to detect the presence or absence of objects, to measure the speed and direction of rotating objects and in other applications.

Infra-red and visible LEDs have rated lifespans of over 100,000 hours, giving opto-switches an effectively infinite operational life. They can switch in four microseconds and operate at high frequency of up to 3,000 counts per second.

These characteristics make transmissive and reflective photo microsensors deservedly popular in office equipment, industrial automation systems, vending machines and home and building automation.

Figure 2: Transmissive slotted photo-interrupter

Figure 2: Transmissive slotted photo-interrupter

Of the newer applications, 3D printers require detection of the position of the print head, and confirmation of correct feeding and movement of the filament; IP cameras need to detect the angle and position of the camera body; and there are countless requirements to detect position in fitness and massaging machines.

Established styles don’t necessarily fully address these new applications. Although they detect most surface textures and colours, they don’t easily detect transparent objects and can be fooled by black items too.

Many have a ‘slotted’ style where the size of the object detected is limited by the width of the slot (Figure 2). They do have a long sensing distance, which can be good, but can also be a drawback as spurious detections can result from objects moving into the background.

New approaches

New technologies are emerging which not only benefit the newer designs, but have advantages for developers of the more established applications too.

Figure 3: Light convergent reflective sensor

Figure 3: Light convergent reflective sensor

Light convergent reflective sensors (Figure 3) detect only objects that are a specific distance from the sensor. They can eliminate background and can detect both specular and diffuse reflecting objects reliably, regardless of their colour or material. They were originally developed for robot cleaning units, which needed to detect and reliably clean floors made of all kinds of materials in any colour.

Particular challenges for traditional optical sensors were detecting floors with glass or black areas. Traditional sensors also struggled to identify downward steps.

Using the new light convergent technology, robot cleaners can be reliably programmed to turn away from ‘cliffs’ and can accurately identify all kinds of floor coverings, implementing the most appropriate cleaning approach for each one.

The same technology is now also being widely applied in printers, allowing them to detect a much wider range of materials including black paper and clear film. Similarly, vending machines can now detect transparent cups, eliminating the need to fit a label to clear glass or plastic items to ensure that they are detected.

Checking the distance

Also new is the introduction of micro displacement sensors. These calculate the distance of the detected object, producing an output voltage proportional to the distance.

Typical devices can detect displacements with a resolution of 10 microns at a distance range of 6.5mm ±1.0mm. They are accurate enough, for example, to detect a double feed in a copier, count the notes in a cash dispenser and detect the amount of paper remaining on a till roll in a mini printer or point-of-sale system.

The first proximity switches were introduced in 1960, and LED type photo electric sensors in the early 1970s.

Although the core principle of pairing an LED light source with a detector remains the same, switches have increased considerably in sophistication and accuracy since then. Light-based proximity sensors continue to develop and it is inevitable that robotics and 3D visualisation will drive further improvements in this technology.

Gabriele Fulco is European product marketing manager for sensors at Omron Electronic Components

Richard Wilson

Source: Electronics Weekly