How to Select and Integrate Multi-Dimensional Safety Systems to Protect Workers from Cobots

Safety is essential when deploying collaborative robots (cobots), autonomous mobile robots (AMRs), and autonomous guided vehicles (AGVs) in factories and logistics facilities. It’s also complex and multi-dimensional.

Machine movements need to be monitored and controlled per International Organization for Standardization (ISO) 13849, International Electrotechnical Commission (IEC) 62061 and IEC 61800-5-2 that provide safety requirements and guidance on the principles for the design and integration of safety-related parts of control systems (SRP/CS).

Ensuring the safe operation of cobots, AMRs, AGVs, and similar equipment often requires the establishment of a layered safety envelope with multiple fields from initial detection and warning of approaching objects to identify when an object breaches a hazardous zone and stops the machine.

A modular safety controller system can add another layer of analysis and protection. Efficient and quick fault analysis can be an important consideration when dealing with protective field interruptions and unexpected tripping of a scanner. That can require a second sensor to monitor the protective field of the primary sensor.

This article starts with a brief refresher on the requirements of ISO 13849, IEC 62061, and IEC 61800-5-2 and a review of the basics of two-dimensional (2D) light detection and ranging (LiDAR) safety laser scanners. It then provides a deeper dive into how layered safety envelopes can be implemented to protect people from cobots, AMRs, AGVs, and similar equipment.

Included is a review of the use and integration of 2D LiDAR sensors and a look at the benefits of combining those sensors with a modular programmable safety controller to provide an additional dimension of safety, plus the use of an event camera to enable fault analysis of unexpected interruptions of protective fields. Exemplary devices from SICK are included.

IEC 61508 is the foundational standard for “Functional Safety of Electrical/Electronic/Programmable Electronic Safety-related Systems (E/E/PE, or E/E/PES)” and applies to all industries. In addition, there are industry and application-specific subsections and variants.

IEC 62061, “Safety of machinery: Functional safety of electrical, electronic and programmable electronic control systems,” is the machinery-specific variant of IEC 61508. IEC 61800-5-2, “Adjustable speed electrical power drive systems – Part 5-2: Safety requirements – Functional,” is also related to IEC 61508 and is a standard for the design and development of adjustable speed drive systems.

ISO 13849 was developed independently and not derived from IEC 61508. Both are concerned with functional safety. IEC 61800-5-2 uses Safety Integrity Levels (SILs) to define safety requirements, while ISO 13849 defines the Required Performance Level (PL_r).

ISO 13849 and IEC 61508 are based on the concept of the probability of dangerous failure per hour (PFH_d). The ISO 13849 functional safety analysis considers three factors: the severity of a possible injury, the frequency or exposure to a hazard, and the potential of limiting the hazard and avoiding harm (Figure 1):

• Severity of injury
  • S1: Slight (normally reversible injury)
  • S2: Serious (normally irreversible or death)
• Frequency and/or exposure to hazard
  • F1: Seldom-to-less-often and/or exposure time is short
  • F2: Frequent-to-continuous and/or exposure time is long
• Possibility of avoiding hazard or limiting harm
  • P1: Possible under specific conditions
  • P2: Hardly possible

Figure 1: Derivation of PL_r levels in ISO 13849 and corresponding SILs in IEC 62061. Both standards are based on the concept of dangerous failure per hour (PFH_d). (Image source: SICK)

How does LiDAR work?

Certification to PL_b according to ISO 13849 is required for the use of 2D LiDAR safety sensors in personal protection applications. The TiM 2D LiDAR sensor family includes models meeting that requirement. 2D LiDAR sensors scan their surroundings using optical time-of-flight (ToF) technology. ToF is implemented by sending laser pulses using a rotating mirror and detecting the reflected light. The longer it takes for the reflected light to arrive back at the sensor, the further away the object.

The time measurement combined with the strength of the returned signal enables the sensor to calculate the position of multiple objects with millimeter accuracy. The resulting picture of the surroundings is updated up to 15 times every second (Figure 2). It can support real-time navigation, orientation, control, and safety functions.

Figure 2: TiM 2D LiDAR sensors use a rotating mirror and laser pulses to create a picture of the surroundings that can be updated up to 15 times every second. (Image source: SICK)

TiM 2D LiDAR sensors detect objects in defined areas (fields) to be monitored. Depending on the model, they have a scanning range of up to 25 m and a working range of up to 270°.

The return pulse data from the laser is processed using high-definition distance measurement (HDDM) or HDDM+ technology. HDDM achieves a very high measurement accuracy at short distances and is suitable for fine positioning in applications like docking. HDDM+ processes edge reflections particularly well, making it best suited for localization and anti-collision applications in dynamic environments.

In both cases, the patented HDDM/HDDM+ multi-pulse technology enables TiM 2D LiDAR sensors to detect the entire scanning range without gaps, ensuring consistent measurement precision, and they can handle different surfaces and remission factors.

Types TiM1xx, TiM3xx, and TiM7xx detect whether objects are in a pre-defined field. Sixteen field sets, each with three preconfigured fields, support quick adaptation during operation (Figure 3). Individual field geometries can be specified, or reference contour fields can be defined for static contour monitoring. Digital filters, masked areas, and response times can also be defined to maximize performance even in the presence of heavy rain, snow, or dust.

Figure 3: Field sets in TiM 2D LiDAR sensors consist of three preconfigured fields. (Image source: SICK)

Models that provide field evaluation data or field evaluation and measurement data are available. Field evaluation sensors only determine the presence of an object, while field evaluation and measurement data can be used to provide an accurate picture of a scanned surface.

In addition to distance data, TiM 2D LiDAR sensors are available that also provide angular data and a received signal strength indicator (RSSI) output. This expanded data set can be especially useful for collision avoidance and navigation for AMRs in changing environments.

Safety LiDARs, adding the first protective layers

The TiM 2D LiDAR family has safety-related variants, the TiM361S (field evaluation) and TiM781S (field evaluation and measurement data output), that meet the requirements of PL_b and can be used for both stationary and mobile applications. They can be used for personal protection in access monitoring for industrial cobots and on mobile platforms like AMRs and AGVs.

• Type TIM361S-2134101, model number 1090608, is suited for indoor use with a detection range of 0.05 to 10 m and HDDM technology.
• Type TIM781S-2174104, model number 1096363, is also suited for indoor use with a detection range of 0.05 to 25 m and HDDM+ technology.

Simplified integration

TiM 2D LiDAR sensors are designed to simplify integration. With an enclosing rating of up to IP67, neither dust nor moisture can enter the housing. They are highly immune to bright ambient lighting up to 80,000 lx. Their rugged design meets the vibration resistance requirements of IEC 60068-2-6 and the shock resistance requirements of IEC 60068-2-27. Their ruggedness can be enhanced when needed using damped mountings of protective plates.

The compact design, light weight, and low power consumption of TiM 2D LiDAR sensors make them well-suited for mobile platforms. The Type TIM361S-2134101 and Type TIM781S-2174104 both weigh only 250 g, have a typical power consumption of 4 W, and measure 60 mm long x 60 mm wide x 86 mm high.

Safety controllers add another layer

LiDAR laser scanners detect hazards and send alerts, while a modular safety controller can add another layer of safety to a protection system. For example, the Flexi Soft safety controller is a modular system that can be connected to various sensors and switching elements, including laser scanners. It’s rated SIL3 according to IEC 61508 and PLe with a PFH_d of 1.07 x 10^-9 according to ISO 13849.

A basic system consists of at least two modules (Figure 4):

1. The CPU0, like model 1043783, is the central logic unit where signals from sensors like LiDAR are analyzed and evaluated, offloading safety analysis from the central machine controller. The output of the CPU0 connects with a higher level machine control, such as a programmable logic controller (PLC), where safety functions are implemented.
2. The XTIO I/O expansion module, such as model 1044125, is required to connect laser scanners to the system. One XTIO I/O expansion module is necessary for every two laser scanners, as each laser scanner uses three switching inputs. The controller can operate up to 12 I/O modules.

Figure 4: The Flexi Soft safety controller system consists of a CPU module (1) and one or more I/O modules (2). (Image source: SICK)

What happened?

An important element in a safety system can be the ability to analyze and understand the root cause of any faults, answering the question, “What happened to cause the safety laser scanner to trigger?” An event camera, EventCam from SICK, is specifically designed to detect and analyze sporadic faults in industrial settings.

EventCam is self-contained with optics, illumination, electronics, and memory and can be integrated into mobile or stationary systems. The cast aluminum housing is IP65-rated and can be mounted in various positions. EventCam can be connected to an automation system like a safety controller or directly to a sensor.

Once an error has been reported, EventCam begins storing single frames or video sequences. The internal ring memory can store up to 240 seconds before and 100 seconds after an event. In high-definition (HD) mode, it can record up to 25 seconds before and 15 seconds after. The video frames per second (fps) rate ranges from 13 to 65, depending on the required resolution.

EventCam can also be useful when commissioning new machines or processes. It can monitor an unsupervised test run like a multi-hour or multi-day continuous test and quickly identify error sources. Multiple EventCams can monitor a single process, providing visual information from several angles at once for a deeper and more thorough analysis of errors (Figure 5).

Figure 5: Multiple EventCams can be synchronized to record a single event from several angles simultaneously. (Image source: SICK)

EventCam is offered in two variants. Model 1102028 has a working range of 0.4 m to 0.6 m and can be suitable for use with stationary cobots with relatively small protective spaces. Model 1093139 has a working range of 0.8 to 6 m and can accommodate larger protective spaces encountered with bigger cobots, AMRs, and AGVs.

Summary

2D LiDAR sensors like the TiM family from SICK can provide the first line of defense in a safety system for cobots, AMRs, AGVs, and similar machines. They provide a series of protective fields to monitor the approach of people. The addition of a safety controller can support intrusion analysis and enhance system performance. Finally, one or more EventCams can monitor the primary 2D LiDAR sensor to help identify the root cause of any sporadic tripping.