Human-machine interface software keeps rails safe
By Ellen Fussell Policastro
Norfolk Southern Railway Company continues a legacy of rail transportation as the fourth largest railroad in the world, operating more than 21,500 route miles in 22 states and in Ontario, Canada. As a firm believer in modern-day automation technology to ensure smooth operation of its rail lines, the company uses supervisory human-machine interface (HMI) software to communicate vital information about its train cars at its Buckeye Hump Yard. Since implementing the system, the company can now provide critical information to the yardmaster, the primary user of the system, with shorter training times.
When the company initially automated the system, operators needed an entire view of the mainframe and mid-range computer systems to effectively manage operations. Then, in the early 1990s, the company installed HMI software, which ran on their Microsoft Windows 3.0 operating system.
Now, all of the information is processed through eight PCs. Implementing a software package for visualization and industrial process control helps users quickly create and deploy customized applications that connect and deliver real-time information. So, they can access applications from mobile devices, thin clients, computer nodes, and over the Internet.
Software manages operations
The software manages operations at the railroad company’s hump yard (a regional gathering point where they classify freight and forward it to its final destinations) in Columbus, Ohio. The process of joining the right train cars to the right engines occurs in three areas: a receiving yard, a classification yard in which railcars are pushed over a hump to various classification tracks, and a forwarding yard.
The panel PCs see use in several control room areas, but the tablet PC works well for this type of environment. “Railyards have a distributed area in which people managing operations need to be able to control things from out in the field,” said Darren Kline from Q-Mation, a distributor who provided support and services for the project.
Norfolk can easily access critical pin puller information using the new HMI displays. “Prior to having a graphical display, the pin pullers had a paper list where they would mark down where all the cars had to be cut,” said Dan Niemiec, principal at Trainyard Tech, an original equipment manufacturer who installed the system.
Today, the pin puller display supplies condensed information about the location of each car, what track it is going to and any special handling codes. And to keep things in sync between the pin puller and the yardmaster, both can access the same information on their HMI screens.
When a train arrives at the hump yard, scanners with antennas collect information from RFID tags on each train car. Alternatively, a field engineer can manually input data from anywhere in the yard into the supervisory system using industrial tablets.
Train yard employees such as the trainmaster, the receiving yardmaster, and the assembly yardmaster can review this information simultaneously since they re-assemble the train cars at the end of the humping process.
Using symbols like asterisks and plus and minus signs, the pin puller display verifies which train cars have arrived at the yard and ensures they match the list of cars Norfolk provides. If there is a discrepancy, the yard personnel can address it right away.
Efficiency, safety key
Increasing efficiency is important because the Buckeye Yard can receive up to 2,000 train cars in any given day. Safety also is key because of the sheer tonnage handled and the dangerous nature of working on the railroad.
The new software has increased the work efficiency of the yardmasters and the trainmasters at Buckeye, making them “more of a point-and-click operation versus having to manage operations manually,” said Tim Forman, trainmaster at Norfolk Southern Railway Company.
The new system has decreased labor costs involved with the repair and upkeep of the system on a day-to-day basis, he said. “We’ve been able to remove a lot of old and antiquated equipment we no longer need due to the upgrades in the current software.”
Buckeye uses a gravity humping technique in which at least 100 train cars weighing up to 12,000 tons are shoved to the top of a hill and then carefully rolled down and directed to 40 classification tracks, where the individual cars are re-sorted and grouped depending on their content and destination.
So much weight rolling downhill can be dangerous for yard workers. Mechanisms similar to automobile disc brakes called retarders slow the train cars down. The software collects and shares information on which retarders are active to keep the workers on the track better informed and safe.
ABOUT THE AUTHOR
Ellen Fussell Policastro is the associate editor of InTech. Her e-mail is firstname.lastname@example.org.
Power plant conversion
HMI features also play an important role in process automation industrial settings, such as the Pyeong-taek thermal power plant, fueled by heavy oil and composed of four power boiler units capable of 350MWs each. In November 2004, the company converted its control system to distributed control system (DCS).
Before the conversion, Pyeong-taek thermal power plant was operated by daily startup and shutdown. Simultaneously, about 17,000 alarms occurred on the human-machine interface (HMI) monitor whenever it started and stopped. The alarms occurred quickly and then faded away before the operators could correctly understand them because they could display only several tens of alarms on the HMI monitor. So an improvement was definitely in order.
The first move was to make the graphic alarm boards for the important alarms on the HMI monitor. The graphic boards consist of six boards: Boiler Main Board 1, Boiler Main Board 2, TBN Main Board, GEN & Common Main Board, Aux alarm & NG alarm Board, and Electric control panel Board.
The second step was to classify the alarms by color and glimmering light.
Third, inspect the causes and countermeasures for the important alarms, produce a book, and distribute and teach the book to the operators.
Help appeared in the window with these choices:
The fourth and final step was to add causes and countermeasures to the graphic alarm board, which allowed operators to see the pop-up window just by moving the mouse point to the alarm window and clicking the left button of the mouse. The company displayed the causes and countermeasures for the alarms on the HMI monitor so now the operators could immediately recognize and correct the alarms. Consequently, they could run the power plant more stably and prevent the unit trip.
SOURCE: Sang-Jin Lee, I&C section manager in the power generation department at Korea Western Power Co., Ltd., “The Complement of DCS Alarm Function,” from ISA’s POWID 2008 conference.
To design successful human-machine interfaces (HMIs), it is necessary to know and understand a suitable design process. However, it is also invaluable to understand how the HMI’s users perceive and remember things, and how they make decisions. It is also important to realize how well they understand the process and the system they use, and to be able to identify beforehand the types of mistakes they are likely to make.
Finally, having a good understanding of operator monitoring activities will be useful in designing interactive monitoring functions.
Visual perception refers to how users see and acquire information relevant to what they are trying to achieve. One of the first things to realize is the legibility of visual information is influenced by a large number of factors such as contrast, font stroke, width, ambient lighting, and the like. From a design point of view, however character height has been used as the main means to affect legibility in HMIs. A suitable character height can be calculated as an angle expressed in minutes or arc. The required angle ranges from 16 to 22 minutes of arc for applications such as process control. The formula to use is:
A is the angle, while H and D are the character height and distance between the operator’s eye and the display, respectively. For example, for a reading distance of 75 cm (typical of table-top display reading) and a character height of 20 minutes, the resulting actual character height in engineering units is:
The formula works fine as long as H and D are expressed in the same unit, whether they are centimeters or inches, etc.
People perceive information then remember what is needed to carry out their tasks and make decisions. The Skill-Rule-Knowledge model of human decision-making, developed by Jens Rasmussen, a researcher in operator modeling from the Riso National Laboratory in Denmark, has become a popular tool to understand operator behavior. Rasmussen’s model divides human behavior as skill-based, rule-based, or knowledge-based behavior.
Skill-based behavior corresponds to the nearly automatic response of an operator handling well-known situations. This produces the best performance in terms of speed, accuracy, and error rate. Rule-based behavior is that in which the operator follows a reasonably well-known process or executes procedures. Knowledge-based behavior is that in which an operator must resort to fundamental knowledge of the process to solve a problem. This behavior tends to be difficult and error-prone.
SOURCE: Human-Machine Interface Design for Process Control Applications, by Jean-Yves Fiset, ISA, 2009.