No more clash of the titans

Online Editor

Neil Sandhu explores the latest collision avoidance technologies

Collisions are a constant hazard in many heavy industrial environments outdoors, especially where large machines frequently work together in confined areas. From ports and terminals to mines, quarries and cement plants, clashes between giant machines, stockpiles, vehicles and personnel must all be avoided. Failure to make adequate provision for collision avoidance on cranes, booms and gantries, as well as manned and unmanned mobile machines, could lead to stoppages, expensive delays, reduced throughput and personal injury.

Domain of the blind

These workplaces are the domain of the blind, presenting challenges for both manned and unmanned machinery to see around every corner and obstacle. Wheel loaders moving backwards and forwards present blind spots to the driver. Objects on the ground may be concealed or obscured by other equipment. Then, there is the environment, whether sea fog at a port, dust from aggregates, snow or rain. Harsh conditions pose exceptional challenges to visibility, as well as to the availability of equipment.

Sensing technologies provide the "eyes" to "see" around the corners and through the dust clouds. Increasingly, with Internet of Things intelligence, they can also feed their data back and integrate with automated software and maintenance systems to provide a wealth of feedback for improved day-to-day operation.

Selecting sensor technologies

Selecting the correct sensing technology to perform reliably in harsh and complex outdoor environments can feel daunting. The risk of false trips is a major concern that could lead equipment specifiers away from optical sensors, even though this choice might be at the sacrifice of more accurate and detailed data.

So, what kinds of "eyes" are best for your site and what criteria should you use to make the choice?

Do you want to measure the distance from the object? Do you need to protect zones or areas of the yard when dangerous machinery is moving within them? Do you need to make a detailed 3D profile of the environment? How accurate do you need to be to detect, measure, or actually identify an object of concern – and how far away is it?

To protect against the worst of the elements, the best technologies may be housed in robust IP67 or IP69-rated enclosures, have protected connectors and may include weather hoods or internal heaters. They may monitor their own operation and send out alerts when they need to be cleaned or replaced.

Let there be light!

Laser distance sensors and lidar are common technologies used for collision avoidance in outdoor industrial environments. They use optical time of flight (ToF) technologies to send invisible infrared light beams to an object and measure the time it takes them to be reflected back to the sensor.

Sick’s laser-based sensors use the company's HDDM+ technology and send out multiple echoes – up to five – to overcome the limitations otherwise caused through bad weather, bright light, dust, smoke or mist. The sensor automatically processes the received signals to filter out irrelevant reflections, such as from water droplets or dust particles, and reliably identifies the actual measurement signals. As a result, the risk of false trips and stoppages is eliminated.

Using a non-contact laser distance sensor, such as the DX1000 is a good choice for longer ranges – up to 1,500m, e.g. mounted on rail mounted or rubber-tyred gantries to avoid collisions, both between cranes and other objects, as well as to measure accurately the gap between two points and determine speeds.

2D lidar sensors

2D lidar sensors scan in a fan-shape around the sensor, creating a plane that can be divided into several detection zones. Mounted on machinery or infrastructure, the best-performing can reach a wide angle of view to the side and behind a scanner, typically up to 275°, although some can scan up to a full 360° around their environment and measure up to ranges of 250-350m.

2D lidar, such as Sick's LMS1000, perform distance sensing and ranging duties to detect moving and stationery objects and output accurate data that can be processed to measure both lengths and widths.

Access to zoned areas can be controlled by creating vertical monitoring areas, for example, to control safe access for people or vehicles. Protection against collisions with over height objects, or with stockpiles, can be achieved by establishing horizontal monitoring areas. Systems can be expanded using RFID tags, GPS or other sensors to identify manned or unmanned mobile vehicles entering a pre-defined area.

3D lidar sensors

3D lidar sensors go further by scanning several measurement levels to map a more detailed profile of the environment and its contours, as well as the shape of objects detected. Sick's AOS lidar solution combines a 2D laser scanner with the firm’s Flexi Soft safety controller to provide a complete system that can help avoid collisions, with sensor self-monitoring to avoid system failures.

Radar sensors

Radar sensors work on similar ToF principles to lidar but they emit electromagnetic radio waves instead of infrared light. Radio waves are not affected by environmental conditions in the same way as light-based technologies, so radar sensors can be the ultimate, super-tough choice for harsh environments and operation during the hours of darkness. The RMS1000 radar sensor can achieve 24-hour detection performance with long-range resolution and distance accuracy.

The key difference is that, while lidar scans in a 2D plane, radar emits its radio waves in a cone shape, which expands over distance. So, radar is less well suited to detecting small objects further away because of its longer wavelengths, especially at longer distances. The RMS1000 performs exceptionally well, for example, to identify a human at 50m, but not so well for smaller objects. However, radar technology may scan a wider environment at longer ranges than an equivalent lidar scanner, but with less accuracy.

Driver assistance systems

For mobile machinery such as wheel loaders, mobile cranes or forklifts, the ability for a driver to monitor blind spots can be critical to avoid collisions with people or equipment. Manoeuvring and reversing are the most frequent causes of accidents. Using sensing technologies to deliver data to a cab-based monitor gives the driver a view of areas that would not be otherwise visible.

Such systems provide both collision avoidance and operator help, but must not be distracting for the operator, for example, by providing an alert for every object encountered around the vehicle. A passive system may not attract the operator's attention in time to avoid an accident. So, proximity detection solutions, such as Sick’s lidar and camera-based systems, include an integrated object detection algorithm and warning strategy that can take both needs into account.

The performance capabilities of the MRS1000 lidar sensor to detect 55,000 measurement points across four layers have made it a popular choice for driver assistance systems. These systems usually comprise one or multiple lidar sensors for environmental sensing around the vehicle, a GPS system for location and speed information, cabling, evaluation unit with operator display in the cab and advanced software for situation-specific warning strategies and visualisation.

Using camera-based monitoring around a vehicle is becoming a popular alternative to lidar, as part of intelligent driver assistance systems. Sick's Visionary B vehicle-mounted collision warning system combines 3D imaging camera technology with an in-cab display to monitor the area behind and at the side of a vehicle. The Visionary B system is based on the company’s robust, high-resolution stereo camera technology that captures 3D data with a single 'snapshot' of light, whether the object is stationary or moving. The driver is provided with a live image, as well as optical and acoustic warning signals, so that objects, people, or other vehicles can be detected around the blind spot.

System monitoring and feedback

Whichever sensor you use for collision avoidance, a wealth of useful information will be provided. Increasingly operators are harnessing this data through gateway systems to manage operations in real-time, analyse incidents and make maintenance decisions through cloud-based monitoring dashboards.

Sick’s Monitoring Box uses a Telematic Data Collector (TDC) connected via Ethernet cable to the sensor, mounted, for example, on a crane. The data is sent securely via LAN, WLAN, or mobile communications, so that the system can be monitored in real time from a smart phone or desktop.

By analysing and comparing data and studying the statistical frequency of near-misses or costly stoppages, hazardous situations can be identified and prevented. Driving routes or crane operations can be optimised to ensure the most efficient operations. Data can also be input into digital maintenance software systems.

Even the harshest environments need not be a barrier to using a wide range of non-contact technologies. With options to integrate with intelligent software and IoT connectivity, it's easy to fit collision avoidance systems into existing machinery, align them and 'teach' them the environment.

Neil Sandhu is Sick’s UK Product Manager for Imaging, Measurement and Ranging. He is also Chair of the UK Industrial Vision Association (UKIVA).

Recent Issues