Support-focused enterprise controls: Object detection for automotive automation

The ability to sense objects and trigger an application is the most overlooked fundamental of control design. Two common forms of object detection are presence detection and movement detection. This is part 2 in a series on standardizing development of programmable logic controller (PLC) programming for controlling discrete manufacturing processes. Link to part 1, below.
By Daniel B. Cardinal December 16, 2014

Object detection is the generic name applied to the use of sensors to identify objects at unique locations in the manufacturing process. The ability to sense objects and trigger an application is the most overlooked fundamental of control design. For control applications, detecting moving objects is integral to setting and resetting application signals. For upper-level system designs, the ability to detect objects is critical to triggering PLC-based system applications. Regardless, many designers do not consider the significant effects various forms of object detection have on control applications. Failure to recognize the importance of this seemingly insignificant fundamental prevents them from grasping the single thread needed to standardize control application circuits. 

Figure 1.1: Presence detection. Courtesy: Daniel Cardinal

Object detection is the primary design focus for triggering and synchronizing all controller-based applications. It involves strategically providing, wiring, and programming event-based sensor circuits to ensure their activations set and reset application variables. The trigger-generated signals are responsible for moving data and activation event-based applications. The physical placement of sensors and mechanical characteristics of mechanical actuators effect trigger designs. To understand these attributes better, designers must study the two base forms of object detection and how mechanical actuators activate sensors to enable trigger circuits. The following definitions help to describe both common forms of object detection:

  • Presence detection: a trigger circuit design that uses one or more sensors to detect the presence of an object at one location.
  • Movement detection: a trigger circuit design that uses multiple discrete sensors or one encoder to detect the presence and absence of an object at two locations.

Regardless of the form of object detection designers select, all designs need physical objects to activate sensors in a deterministic fashion. All control system designs have already adapted to the type and arrangement of sensors provided as part of a machine’s mechanical design. The prearrangement of sensors dictates how objects need to actuate them. The sensors selected always match the physical characteristics of mechanical actuators. 

4 object detection functions

The following names describe common sensor functions used to detect objects:

  1. Part present sensor: a sensing device strategically placed to detect the presence of an object.
  2. Entering sensor: a sensing device strategically placed to detect an object entering a process station.
  3. In-position sensor: a sensing device strategically placed to detect the arrival of an object at a process station.
  4. Exiting sensor: a sensing device strategically placed to detect an object exiting a process station.

It is imperative that designers recognize the various circuits used to enable object detection triggers. A presence detection circuit generates an application trigger after all circuit sensors activate. The enabled trigger represents the static presence of an object, whereas the disabled trigger represents the absence of the object. A movement detection circuit enables two momentary triggers when sensors identify an object passing through a station. This circuit design enables a signal when the object stops at a station. When the signal is on, it represents the presence of the object while it is at or between both sensor positions. When the signal is off, it denotes the departed absence of the object. 

5 differences: presence, movement

There are five distinguishing design attributes that differentiate presence detection from movement detection.

  1. Total number of triggers a circuit produces.
  2. Minimum number of circuit sensors needed to enable a single trigger.
  3. Length of the mechanical actuator with respect to the length of an object.
  4. Sensor’s on to off signal activation sequences.
  5. Complexity of circuit designs that affect the retentive or nonretentive nature of circuit signals.

For detailed comparisons, it is useful to examine both forms of object detection circuits using two sensors.

Object detection circuit designs produce application triggers when mechanical actuators activate sensors. Reliable movement detection designs rely on smaller length actuators, thus ensuring shorter sensor activation times. Presence detection designs rely on long actuators, thereby ensuring extended sensor activation times. Both types of object detection circuits need sensors to deactivate between each object. This is an important design consideration when arriving objects travel close or butted together.

Simply put, presence detection senses an object at one physical position, whereas movement detection senses an object at two positions. The difference is important when understanding the system-related benefits of movement detection. This includes the design details associated with sensor placements, mechanical sensor actuators, sensor activation timings, trigger circuits, and trigger reliability. Providing designers with this information will enable applications to obtain the beneficial effects of movement detection triggers. 

Presence detection

Presence detection circuits typically use more than one sensor to identify the presence of objects as they enter or stop at stations. Most designs need two sensors where the initial sensor activates as an object enters, and then the other activates as an object stops in position. The second sensor acts as a filter because it eliminates the intermittent oscillating effects of the first sensor. This oscillation effect is the primary reason control applications do not use presence detection to activate event-based control applications.

Figure 1.1 shows the sensor and actuator arrangements for both the part and carrier-actuated forms of presence detection. The part-actuated form uses the part to activate both sensors, whereas the carrier version uses a long mechanical actuator attached to the carrier to activate and maintain both sensors. Some designs rely on two carrier actuators.Many manufacturers do not allow sensors to come in physical contact with the product. Carrier actuation is one way of presence detection.

Figure 1.2: Empty-carrier presence detection. Courtesy: Daniel Cardinal

Figure 1.2 shows the physical relationship between a part present sensor and two object sensors that detect a carrier’s actuator. Since the object arriving at the station is a carrier, a presence detection circuit includes the enabled state of the part present sensor. The disabled state of this signal blocks the circuit from enabling a normal part in-position trigger and alternately activates an empty carrier trigger. One of the two triggers will always fire the moment both object sensors activate.

When objects are traveling close together, presence detection circuits rely on two physical object related dimensions. First, the length of an object’s actuator must be greater than the distance between both sensors. This dimensional aspect of the mechanical design ensures an actuator has both sensors activated at the same time. Second, the distance between two consecutive carrier actuators must be greater than the length of the object’s actuator. This dimensional aspect ensures both sensors are off at the same time between objects. Deactivation guarantees both sensors will separately reactivate for the next object to enter the station. Figure 1.3 shows the critical actuator and sensor spacing dimensions.

Figure 1.3: Critical presence detection sensor dimensions. Courtesy: Daniel Cardinal

  • L > S + C: The minimum sensor actuator length (L) must be greater than the spacing distance (S) between the two sensors plus a coast distance constant (C). The coast constant is the distance the actuator will travel when sensor and control power fails and before mechanical movement stops. This will ensure an applied controller always will see both sensors on after power restoration.
  • A > S + F: The distance between the sensor actuators (A) must be greater than the spacing distance (S) between the two sensors plus the field width of a sensor (F).

Figure 1.4: Presence detection sensor activation timing-diagram. Courtesy: Daniel Cardinal

Figure 1.4 shows a timing diagram for two presence detection sensors activating for two consecutive objects entering and exiting a station. When the objects enter the station, they sequentially activate and deactivate both sensors. The subsequent numeric sequence describes the numbered steps shown in the timing diagram.

  1. Object entering sensor on: The first sensor activates, signifying an object is entering the station while the second object in-position sensor remains deactivated.
  2. Object in-position sensor on: The second sensor activates, signifying an object is in position at the station while the first object entering sensor remains activated.
  3. Object entering sensor off: The first object entering sensor deactivates, signifying an object is exiting the station while the second in-position sensor remains activated.
  4. Object in-position sensor off: The second sensor deactivates, signifying an object has exited the station while the first object entering sensor remains deactivated.

Figure 1.5: Hard-coded presence detection trigger for machine controller. Courtesy: Daniel Cardinal

Figure 1.5 shows how control system designers use hard-coded circuit triggers in support of event-based machine controller applications. The design uses a standard method for individually wiring sensor contacts to the machine controller’s input module. The input module then separately sends discrete signals to the machine controller’s control application. Control system designers use a ladder logic circuit that examines input signals from each sensor. When the hard-coded circuit recognizes that both inputs are on, it enables a discrete trigger signal.

Figure 1.6: Presence detection trigger timing diagram. Courtesy: Daniel Cardinal

Figure 1.6 shows a timing diagram for both presence detection sensors and a circuit produced in-position trigger. Specifically, the signal timing shows how the trigger remains enabled while an object moves through a station. This extended time makes presence detection triggering ideally suited for controllers that cannot quickly identify and react to fast-changing sensor inputs. If sensor inputs do not stay on long enough to activate controller applications, designers integrate timers with their trigger circuits to ensure triggers stay on for a predetermined minimum amount of time.

The use of presence detection circuits is the reason control systems have many undesirable design traits. Designers usually refer to these traits as idiosyncrasies or anomalies. In most cases, undesirable system traits directly result from designs using presence detection circuits, and the associated design needs to add redundant sensors. Presence detection circuits are unreliable because they are more likely to produce unwanted extra triggers. Alternatively, they may not activate, causing missing triggers. These types of events have an induced effect on control system applications that manifest themselves in many problematic ways.

Presence detection circuits activate to produce unwanted triggers when power failures occur, or when an object prematurely stops near an in-position sensor. Power failures cause extra triggers when power restoration reactivates circuit triggers. Objects that prematurely stop near a sensor’s activation point make the trigger circuit susceptible to physical vibrations or temperature conditions that cause sensors to erroneously oscillate input signals. False triggers occur when associated circuits react to oscillating sensor input signals.

False sensor activations and added redundant sensors are the biggest reason why presence detection triggers misfire. False activation causes designers to add time delay circuit filters to ensure sensors remain off and then go on for a predetermined amount of time before they allow input signals to re-enable a circuit’s trigger. These time-delayed trigger circuits are susceptible to the physical spacing delays between moving objects. If objects arrive too closely spaced together, and if deactivation delay times are too long, circuits fail to enable triggers. The failure of added redundant sensors to deactivate or activate is the cause of most missed triggers. 

Movement detection

Movement detection circuits use two sensors that sequentially identify the presence and absence of objects at two unique station positions. For automated stations, the positions are typically in-position and exiting. Movement detection circuits always use at least two sensors at two momentary actuation positions.

Figure 2.1: Movement detection. Courtesy: Daniel Cardinal

Figure 2.1 shows the sensor and actuator arrangements for the part and carrier-actuated forms of movement detection. The part-actuated form uses the small part protrusion to activate both sensors, whereas the carrier-actuated form uses the short actuator attached to the carrier to toggle each sensor on and off. For comparison purposes, the carrier-actuated form of movement detection is described below.

Figure 2.2: Empty carrier movement detection. Courtesy: Daniel Cardinal

Figure 2.2 shows the physical arrangement of a part present sensor and two object sensors. For this example, an empty carrier trigger fires when the carrier’s mechanical actuator activates the station’s object in-position or object exiting sensor. Since the object identified is a carrier, a movement detection circuit can include the part present sensor. The absence of a signal from this sensor blocks the circuit from enabling a normal position trigger and alternately activates an empty carrier trigger. Based on the position of the part present sensor, circuit designs can enable an empty carrier trigger at both positions.

Control applications always use some form of movement detection to generate event triggers. The next article on control system triggers describes the various forms of movement detection and the many electrical, mechanical, and logical design differences. The worst-case circuit design scenario occurs when objects travel close or rammed together. For this scenario, all movement detection circuits must rely on the distance between sensors to be greater than the length of the actuator. Figure 2.3 shows these critical actuator lengths and sensor-spacing dimensions.

Figure 2.3: Critical movement detection sensor dimensions. Courtesy: Daniel Cardinal

  • L > C: The sensor actuator length (L) must be greater than the distance the object will coast (C) when sensor control power fails and before mechanical movement stops. This will ensure machine controller applications always will see sensors on before power fails or after power restoration.
  • S > L: The spacing distance (S) between the two sensors must be greater than the length of the sensor actuator (L).
  • A > S + F: The distance between sensor actuators (A) must be greater than the spacing distance (S) between the two sensors, plus the field width of one sensor (F).

Figure 2.4: Movement detection sensor activation timing-diagram. Courtesy: Daniel Cardinal

Figure 2.4 shows a timing diagram for two movement detection sensors activated by two consecutive objects entering and exiting a station. When the two objects enter the automated station, they sequentially activate and deactivate both sensors. The subsequent numeric sequence describes the numbered steps shown in the timing diagram.

1. Object in-position sensor on: The first sensor activates, signifying an object is arriving in position at a station while the second object exiting sensor remains deactivated.

2. Object in-position sensor off: The first sensor deactivates, signifying the arrival or beginning departure of an object while the second object exiting sensor remains deactivated.

3. Object exiting sensor on: The second sensor activates, signifying an object is exiting the station while the first in-position sensor remains deactivated.

4. Object exiting sensor off: The second sensor deactivates, signifying an object has exited the station while the first in-position sensor remains deactivated.

Movement detection triggering characteristics are important to event-based applications, whereas presence detection or the sensed stationary presence of objects is integral to state-based applications. As a result, control system designers use machine controller circuits to trigger event-based applications for many reasons. Their circuits simply include the movement detection sensors to control the behavior machines. Furthermore, machine controller application environments automatically support the design characteristics of movement detection circuits. Machine controllers have the inherent ability to detect and process fast-changing sensor input signals. Machine controllers also have the ability to set and retain signal states that memorize the presence of objects when controllers are shut off or experience unexpected power interruptions.

Figure 2.5: Hard-coded movement detection trigger for machine controller. Courtesy: Daniel Cardinal

Figure 2.5 shows how designers program two movement detection circuits for a machine controller’s event-based application. Designers understand that each trigger will remain on for one scan of the entire machine controller program. One program scan is long enough for every controller-resident application to examine and react to the trigger. The illustration also shows how the hard-coded circuits produce simple application triggers. The basic idea is to wait for each sensor to produce an application trigger at two unique station positions.

Figure 2.6: Movement detection trigger timing diagram. Courtesy: Daniel Cardinal

Figure 2.6 shows a signal timing diagram for the movement detection sensors, and the associated circuit produced in-position and exiting triggers. The use of movement detection circuits is the reason most system designs have few or no operational idiosyncrasies and unwanted anomalies. This results from the reliable nature of two-step, two-trigger sequence. The term "reliable" in this context means the trigger circuits produce fewer unwanted extra triggers or not activate to cause missed triggers.

Movement detection circuits are less susceptible to power failures. This is because the power on re-toggling of any one sensor does not cause extra triggers. Applied movement detection circuits only create unnecessary triggers when both sensors toggle in the described two-step sequence. Since most designs are an integral part of control applications, the design creates extra triggers when a sensed object moves back and forth in the prescribed sequence.

Movement detection circuits are also less prone to fail and produce missed triggers. This is because integrated designs do not release parts if sensors fail to toggle during the normal cycle of a machine. Control applications detect sensor failures by verifying that sensor input signals properly go on and/or off while the machine is cycling. The only way triggers fail to fire is when a power interruption occurs, and the moving object coasts through a sensor’s deactivated field. Since sensors are critical to normal sequences, machine designers minimize these types of problems by providing power off safety brakes and long sensor actuators. Both logical and physical design attributes are inherent aspects of good control system and mechanical machine designs.

Daniel B. Cardinal graduated from Michigan Technological University as an Electrical Engineer. Mr. Cardinal has over thirty years of experience designing control systems and working as a Controls Integrator. Courtesy: Daniel CardinalThe next article on control system triggers describes the various forms of movement detection and the many electrical, mechanical, and logical design differences.

– Daniel B. Cardinal graduated from Michigan Technological University as an electrical engineer. Mr. Cardinal has over 30 years of experience designing control systems and working as a controls integrator. In the early 1980s he was a controls supervisor for one of Europe’s largest machine tool suppliers, helping this company establish a presence in the United States. Later, he was a co-owner of Control Systems Associates Inc., a company that specialized in integrated system designs; edited by Joy Chang, digital project manager, Control Engineering, jchang@cfemedia.com.

See the first article in this series below.