Reduce human errors: Get out of the typing pool
Engineering and IT Insight: Error-proofing can include adding automation to processes that are dangerous, dirty, exacting, or repetitive. Manual entry of signatures, dates, and times is a candidate. Use keyboard wedges to scan predefined reason codes, enter user IDs, and provide biometric identification. Convert voice to text to reduce errors and speed manual data entry.
Manufacturing companies have long realized that the greatest source of error in manufacturing is human error, and it is a large part of the reason that automation is applied to any process that is dangerous, dirty, exacting, or repetitive. However, in many systems, there is another source of human error that is on the rise. This source is manual recording of data and comments, and in the increasingly regulated manufacturing environment, signatures, date, and times are becoming more common. Comments must also be added to production records to record the reason for discrepancies, providing data for productivity improvements and 6-sigma projects.
Manually typed data, including signatures, values, and comments, is inherently error-prone. Even best-in-class typists have error rates that are about 4 sigma processes (99.4% correct, or less than 1 error per 100 words before corrections). Production operators are not best-in-class typists; typing skills are usually not part of their required skill set or included in their training plans.
Consider your own error rate when entering a phone number, entering a date, or even writing a simple comment on an unexpected event. If you have error rates of 3 sigma (about 7 errors per 100 words before correction), then you have excellent typing skills.
How can we design 6-sigma processes when the manual data entry part of the process has error rates about 3 sigma? This problem has plagued the regulated industries for a long time. Manufacturing execution systems (MES) have replaced written entries with typed entries and have removed the need to record dates and times. Unfortunately, poor penmanship has been replaced by mistyped data. Operators must still enter user IDs and passwords for electronic signatures and must enter comments and observations on production events. User IDs and passwords have a built-in error check, but even entering these short strings produces enough errors that industrial users turn off bad password lock-out checks.
The answer to reducing errors is to remove the need for typing in as many places as possible, and there are some simple techniques to use. A primary method to reduce typing is to use a “keyboard wedge.” A keyboard wedge is a device that that simulates keyboard data. It reads magnetic strips, bar codes, radio frequency identification (RFID) tags, smart-card codes, or other input, which is then routed to the keyboard buffer so it appears as if it was typed in. The term “wedge” comes from the fact that the device usually wedges between the keyboard and the system. The growth of USB ports means that any system can support multiple keyboard wedge devices.
Most companies’ employee cards now include an RFID chip for use with badge readers. A keyboard wedge that reads the employee badge can be used to enter the ID for an electronic signature. This method cuts by more than half the time to enter an electronic signature. Employees don’t have to remember their ID and then type it correctly, so errors and retries are eliminated. If your application supports hot keys (such as F5) that move the cursor to the User ID field, then even more time is saved and errors reduced.
Bar code scanner keyboard wedges can also be used for entry of predefined comments and observations. Most production lines have a small number of problems that cause the majority of the production discrepancies. A bar code scanner and a printed sheet with common problems encoded in bar codes can eliminate the need for manual typing of observations and comments. This reduces the time to enter comments and eliminates errors.
Fingerprint scanners and facial recognition can be used to eliminate the need for any keyboard interaction for electronic signatures. Both systems usually directly connect to a system’s underlying operating system. Fingerprint scanners provide a biometric method for identifying a person, and when combined with an RFID-enabled badge, meet the requirements for an electronic signature. Fingerprint readers are available on laptops or integrated with keyboards. Not all environments are suitable for fingerprint readers, but an alternative is facial recognition. An inexpensive webcam combined with the appropriate software also can be used for electronic signatures. Facial recognition systems have very low false-positive error rates, even in environments where operators must wear partial facial protection, such as hair nets and beard nets. Facial recognition systems also have the capability of recording a picture of the person electronically signing, providing undisputable evidence of the electronic signature.
In the future, two methods are emerging that could further eliminate manual errors on typing. These are voice recognition and gestures. While voice recognition systems have made fantastic advances in recent years, their error rates in office environments for free text entry are still in the same range as manually typed data. The accuracy rate is further reduced in production environments because they are noisier than offices. Voice recognition can be effectively used to replace simple typed commands, usually from a limited set of valid responses. Voice recognition can be especially useful in environments where an operator’s hands are unavailable for typing, such as in warehouses, gloved environments, and driving situations.
Gestures are becoming popular in video games but have not yet made the transition to commercial use. Anyone who has used a Microsoft Kinect or watched a Hollywood high-tech movie can see computers being controlled by gestures. It’s not clear yet what the error rates will be on gesture-defined input, but it is hard to image that it will be lower than manually typed data.
Reducing manual data entry errors is possible. You can reduce or even eliminate the inherently 3-sigma process of typing on your systems by using a combination of alternate entry devices. Use keyboard wedges to scan predefined reason codes, enter user IDs, and provide biometric identification. Convert voice to text to reduce errors and speed up manual data entry tasks. All of these devices will get your operational staff out of the typing pool and back to value-added production activities.
- Dennis Brandl is president of BR&L Consulting in Cary, N.C., www.brlconsulting.com. His firm focuses on manufacturing IT. Contact him at dbrandl(at)brlconsulting.com; Edited by Mark T. Hoske, content manager CFE Media, Control Engineering.
- Read more Engineering and IT Insight columns: atop www.controleng.com search “Brandl”
- Search “HMI design” to gain more advice
- http://controleng.com/webcast has a webcast on HMI design
Case Study Database
Get more exposure for your case study by uploading it to the Plant Engineering case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.
2012 Salary Survey
In a year when manufacturing continued to lead the economic rebound, it makes sense that plant manager bonuses rebounded. Plant Engineering’s annual Salary Survey shows both wages and bonuses rose in 2012 after a retreat the year before.
Average salary across all job titles for plant floor management rose 3.5% to $95,446, and bonus compensation jumped to $15,162, a 4.2% increase from the 2010 level and double the 2011 total, which showed a sharp drop in bonus.