Monday 13 August 2012

Applied Ergonomics


Introduction:
The summary of 23 highly useful ‘‘lessons learned’’ in applying ergonomics to the analysis and design of systems, and supporting documentation is presented here. This will be useful to ergonomists in various ways, including applying ergonomics to system analysis and design and ‘‘selling’’ ergonomics programs and projects to clients, or to one’s own management. As used herein, a ‘‘system’’ can be as simple as a single person with a tool to as complex as a multinational corporation.

Summary of the ‘‘lessons learned’’, described herein
  1. The science and practice of ergonomics is the same throughout the world
The emphasis on particular facets or applications of ergonomics may differ from country to country, but the ergonomics professionals are concerned with design of the interfaces between humans and the other system components for improving health, safety, comfort and productivity, including quality, and reducing design induced human error. As practiced universally, the over-all goal of human factor professionals is to improve the quality of human life.
  1. Its unique technology defines and scopes ergonomics as a discipline
As a practice, ergonomists around the world apply human–system interface
technology (HSIT) to the design or modification of systems to enhance safety,
health, comfort, and performance, including productivity and quality. These goals are achieved through applying HSIT to the analysis, design, test and evaluation,
standardization, and control of systems. It is the development and application of HSIT that both identifies ergonomics as a unique, scientifically based discipline and defines its current scope. HSIT, in turn, has at least five identifiable major components
Ø  Human–machine interface technology or hardware ergonomics
Ø  Human–environment interface technology or environmental ergonomics
Ø  Human–software interface technology or cognitive ergonomics
Ø  Human–job interface technology or work design ergonomics
Ø  Human–organization interface technology, or macro ergonomics
  1. Ergonomics technology can be applied to any system, product, or built
environment
All socio-technical systems involve the interface of humans with other system components. All human–system interfaces require consideration of the same human
capabilities, limitations and other characteristics. All require the application of the same scientifically based HSIT.
  1. Ergonomics is not simply a lay person’s ‘‘common sense’’; professional expertise
is required
Unfortunately, from the author’s experience, there are many persons who call themselves ergonomists or human factors specialists who have virtually no education and training in HSIT and its application. The development of professional certification programs in ergonomics—such as the Euro Ergonomist (CREE) and the IEA endorsed Board of Certification in Professional Ergonomics (BCPE) programs—serves to help organizations identify persons who have the necessary professional education, training, and experience in HSIT.
  1. Good ergonomics projects typically give a direct cost–benefit of from 1 to 2, to 1 to 10+,  with a typical pay-back period of 6–24 months
Of particular note is the fact that life cycle cost savings can be of even greater importance than the initial direct savings.
  1. Effective ergonomics programs on large system development projects take only 1% of   the engineering design budget
  2. The earlier ergonomics is applied in design, the cheaper the cost and greater the
Benefit
The earlier there is professional ergonomics participation in the design team’s work, the less costly is the effort. From the author’s experience, the ergonomics portion of the engineering budget increases when ergonomists are brought in late in the project because serious human–system interface problems have surfaced that require major retrofits in order to correct them. The same trend in costs also applies to software development.The cost of catching and solving ergonomic usability design problems early in the software design process costs about one-fourth of what the same changes made late would cost.
  1. The language of business is money
Managers have to justify any expenditure in terms of the cost–benefit ratio—how the project will affect the bottom line. Accordingly, we must express ergonomic project proposals in financial terms. Costs include such things as the cost of personnel, equipment, and materials, any reduced productivity or loss of sales during implementation, and overhead costs. Benefits include personnel savings; reductions in scrap, rejects, production parts, material, and overhead; and increases in output, sales, or company stock value. It is important for us to conduct a cost–benefit analysis of the various feasible ergonomic solutions, and be prepared to present our analyses to management in a clear and precise manner.

  1. Costs and benefits of ergonomics projects must be measured

We must measure the actual costs and benefits of our projects to show the actual value added of ergonomics—and share that information with others. It is through actual documentation of our value added that we gain credibility with decision-makers and get new opportunities to apply our knowledge.
  1. The trade-off diamond is a useful tool for evaluating interventions
Imagine a diamond in which the shape can change to lengthen or shorten one of the four points, and that each of the points represents a basic intervention strategy: (a) personnel selection, (b) training, (c) human–system interface design, and (d) job performance aids. In general, as one of these points gets ‘‘lengthened’’ or implemented, the need for the other strategy points diminish. Thus, if one better designs the human–system interfaces, the need for additional training or hiring people with a higher skill level diminishes.
  1. Ergonomic improvements to reduce accidents and work-related musculoskeletal disorders usually improve productivity—and vice versa
  2. Pick the ‘‘low hanging fruit’’ first
Obtain quick results from micro-ergonomic interventions to gain credibility with management. Then you are in a position to make macro ergonomic improvements to work systems. This typically is accomplished by selecting those obvious deficiencies that readily can be fixed and show positive results quickly, such as excessive lifting and awkward postures.
13. Look for the simple, economic solution first
      14. Less tangible benefits from ergonomic improvements also can have a significant
            economic impact
Included can be improved employee satisfaction and commitment, which leads to ‘‘good citizenship’’ behaviour (e.g., willingness to work overtime to get the job done, and better collaboration with others or other departments) and an improved corporate image, which can result in less governmental scrutiny and better community relations can have a positive financial impact.
      15. Employee ergonomics training is important to safety and productivity
      16. Real management commitment essential: deeds, not just words
Real commitment gets reflected by such things as hiring qualified ergonomics and safety professionals, providing ergonomics training to all employees, implementing and supporting ergonomics and safety committees, funding appropriate risk analyses, and funding follow-up corrective actions when they can be justified from a cost–benefit standpoint.
      17. Collaboration invariably works better than confrontation
When the ergonomist respects the ownership and design expertise of the design engineer for his/her part of the project, and is collaboratively supportive of that engineer, that engineer is likely to be far more open to, and accepting of, the ergonomist’s input.              
     18.  Ergonomists can be effective system integrators on system development projects
Because ergonomists get involved in the ergonomic design aspects of all system components and subsystems, they are likely to know more about the entire system than any other single engineer. They thus can sometimes see possibilities that others miss.
    19.  OSHA guidelines really do work
In every case where the OSHA guidelines for a professional ergonomics program have been implemented, the organization has experienced significant reductions in lost time accidents and injuries (and often, improved productivity). On the other hand, in those cases where major elements of the OSHA guidelines have not been implemented, the ergonomics and safety programs was found to be inadequate and the accident and injury rate to be unnecessarily high and, often, productivity to be sub-optimal. This is not surprising in that the OSHA guidelines are based on extensive research on what actually works and does not work.
20. Participatory ergonomics is a proven methodology for ensuring that the derived    
      benefits of a macro ergonomic intervention will last
Employees best know the problems with their jobs and which ergonomic alternatives will be most satisfying to them. When, through participatory ergonomics, they are involved in the process, they are likely to ‘‘buy in’’ to the work system changes.
21. True macro ergonomics interventions typically achieve a 50–90% improvement     
      in one or more work system effectiveness criteria
The use of a macro ergonomic approach for implementing TQM at L.L. Bean, a US manufacturer and mail order catalogue distributor of high quality clothing (Rooney et al., 1993). Using methods similar to those for Imada’s petroleum distribution company intervention, but with TQM as the primary objective, over a 70% reduction in lost time accidents and injuries was achieved within a 2-year period in both the production and distribution divisions of the company.
     22. Human-centered design of products and systems is the sure way to success
     23. Ride the coattails of the latest management fad
One way to sell a macro ergonomics intervention is to integrate it with whatever is the hot management program or fad at the time.

Conclusion:
This issue states that Ergonomists have a responsibility to document the cost–benefit analysis of a proposed ergonomics project and advertise those results to management, government decision-makers, and the public in general. It is only through these efforts that they can raise the consciousness of others to the value of ergonomics and gain their support. To achieve the potential of ergonomics, the scientific literature need to be translated into practical ‘‘how to’’ guidelines and specifications for practical engineering design use. Much research is needed to determine the outcome predictability of our interventions. At present, at best, outcomes can be predicted within a broad range only. Given the rapid advancements in technology and the profound challenges those changes will provide for this discipline, research is needed to ensure that the knowledge base of Ergonomists is adequate to meet those changes.

CMOS Image Sensors and Camera-on-a-Chip for Low-Light Level Biomedical Applications


CMOS Image Sensors and Camera-on-a-Chip for Low-Light Level Biomedical Applications

Abstract:
With the advances in deep submicron CMOS technologies, CMOS-based active-pixel sensors (APS) have become a practical alternative to charge-coupled devices (CCD) imaging technology. Key advantages of CMOS image sensors are that they are fabricated in standard CMOS technologies, which allow full integration of the image sensor along with the analog and digital processing and control circuits on the same chip and that they are of low cost. Since there is a practical limit on the minimum pixel size (4~5 μm), then CMOS technology scaling can allow for an increased number of transistors to be integrated into the pixel. This truly shows the potential of CMOS technology in imaging applications, especially for high-speed applications. This work discusses various active-pixel sensors (APS) and shows the feasibility of using the DC-level to increase the sensitivity of the pixel for low-level light applications. Avalanche-photodiodes (APDs) are described, in addition to a discussion of the breakdown mechanism and micro plasma in avalanche breakdown for single photon APDs.

Introduction:
Emerging optical molecular imaging systems have had a revolutionary impact on medicine, bio defense and environmental testing through techniques such as DNA sequencing, protein detection, and evaluation of animal models of human cancer. The most sensitive optical detection system in use is the photomultiplier tube (PMT) but not preferred as they are costly.
Two alternative image sensors that can be used for optical molecular imaging systems are charge-coupled devices (CCDs) and CMOS imagers. CCDs must remain cooled in order to increase their sensitivity to low-level light for biomedical applications. Also, CMOS image sensors consume less power, operate at higher speeds, and offer much higher levels of integration. The advances in deep submicron CMOS technologies have made CMOS image sensors a practical alternative to the long dominating CCD imaging technology. One of the main advantages of CMOS image sensors is that they are fabricated in standard CMOS technologies, which allow full integration of the image sensor along with the processing and control circuits on the same chip at a low cost.  A CMOS camera-on-chip system leads to reduction
in power consumption, cost and sensor size and allows for integration of new sensor functionalities. Since there is a practical limit on the minimum pixel size (4~5 μm), then CMOS technology scaling can allow for an increased number of transistors to be integrated into the pixel. Since digital transistors benefit more from CMOS scaling properties, digital pixel sensors (DPS) have become very attractive. A DPS integrates an ADC in each pixel, resulting in massively parallel readout and conversion that can allow very high speed operation, where digital data is read out of each pixel. The high speed readout makes CMOS image sensors suitable for very high-resolution imagers (multi-megapixels) particularly video applications.


Fig. 1 shows a block diagram of the CMOS imager setup, where the CMOS imager is controlled by the FPGA board

Different PIXEL Structures:

Different pixel structures are used for CMOS imagers. Each pixel structure has its advantages and is can be suitable for specific applications. In the following sub-sections, some common CMOS pixel structures are presented, and their applicability to low-light-level applications are discussed.

1. Passive-, Active- and Digital-Pixel Sensors
Passive pixel sensor (PPS) is the earliest and most simple CMOS pixel structure. In PPS, each pixel consists of a photodiode and a row-select transistor. The PPS has only one transistor per pixel, and thus it has the highest FF. The active pixel sensor (APS) is the most popular sensor.

2. CMOS pixel structure DC Level Mode
APS Active pixel sensors, in general, have an output with low signal-to-noise ratio (SNR) for low-levels of light. One way to increase the sensitivity of the APS is to increase its photodiode’s size. This solution, however, will decrease the resolution of the imager. Measurement results show that the DC level of the output can detect light levels which are two decades or less compared to the swing of the same pixel. For the light power at the low levels of, the SNR of the DC level stands well above the conventional APS.

3. Pixels with Avalanche Photodiode
All of the above pixel structures operate by integrating the photocurrent. In applications where the signal is changing very fast, short integration times are necessary to obtain the desired temporal resolution. However, detection of lower levels of light requires the small photocurrent to be integrated during longer integration times. These above approaches cannot serve the applications that require sensitivity and fast response at the same time. A regular p+/n-well diode was fabricated in standard CMOS technology as a p+ region implanted within an n-well region. In this diode, the breakdown current will not flow uniformly across the area of the p+region. The breakdown region of such diode will be at its edge. This is due to the higher peak electric field caused by the narrower depletion region at the corners of the diode junction. As the reverse bias increases, the electric field at the perimeter will reach the onset of avalanche first, and the current will flow there. However, in APDs, the breakdown region should be spread over the area of the diode and not at its corners. We have made this possible by creating a p-type guard ring around the p+ active area of the APD. To create the guard ring, an n-well region is placed within the p+ region, which is against the conventional design rules of the standard CMOS. The width of the guard ring is 3 μm, with a depth of approximately 0.5 μm. It is shown that the designed device has excellent avalanche characteristics. However, it should be noted that standard CMOS technology is targeted for digital and analog applications and not optical imaging devices.

APD Breakdown and Micro plasma

The proper APD layout ensures that the maximum electric field happens across the active area of the APD, rather than at its edges or corners. Impact ionization requires an electric field of at least E = 300 kV/cm. In a reverse-biased diode, the peak is located at the metallurgic junction. When the reverse bias is just above the breakdown voltage, a narrow strip of high electric field region around the metallurgic junction is where impact ionization occurs, rather than in the entire depletion region. In this narrow region, any device imperfection can cause a local disturbance of the electrical field that can lead to a reduction of the breakdown voltage to a value below the breakdown voltage of the surrounding uniform junction. These tiny spots will be the site of the localized avalanche breakdown of the device. This breakdown condition is generally regarded as being a solid-state analogy of gas discharge plasma, and it is called micro plasma. The micro plasma can occur at threading dislocations, metal-rich precipitates, diffusion induced stacking faults, dopant impurity dislocations, diffusion voids and cracks or mechanical damage. At the onset of avalanching, micro plasmas switch on and off randomly, producing current pulses of constant height. The micro plasma is on for an increasing fraction of the time as the voltage increases until it becomes quiescent. The current carried by micro plasma is limited by heating, spreading resistance and space charge effects.

 CMOS Imager Design

In this camera-on-a-chip, the row and column scanners are used instead of decoder circuits in order to
reduce the control lines coming into the chip. Also, only one input clock is needed to control the row and column circuitry. The multiplexed output is buffered by an on-chip op-amp, which provides the chip’s analog output that is only used for testing and comparison purposes. The analog voltage is routed to the sample-and-hold (S/H) circuit and ADC. The ADC used is a 6-bit dual-slope integrating topology clocked with a 1 MHz clock that is provided by the FPGA. A maximum of 64 clock-cycles is required to complete the conversion, which results in a frame-rate of60 frames/s. The chip provides both parallel and serial outputs from the ADC for testing purposes. All elements, including the state-machine, counters and buffers are implemented on chip. The CMOS imager is tested on an optical table with a 25 mm diameter achromatic lens that has an effective focal length of 15 mm.  The Altera DE2-70 FPGA board is used for controlling the CMOS imager and all of the FPGA code is written in Verilog. The tested targets are
kept 44 cm away from the imager and have a height of about 2 cm.

Conclusion

This paper demonstrates the great potential of CMOS technology to be used in biomedical imaging applications. The DC level APS is an excellent choice for low-level light applications. It uses the same pixel structure as the APS. For applications that require both sensitivity and fast response, APDs should be selected. The APDs with active peripheral circuitry offer the highest speed at the cost of larger pixel sizes. However, by incorporating the advantage of small transistor sizes of modern CMOS technologies, APDs with peripheral circuitry become an excellent choice for achieving both high speed and high sensitivity performance. Also, single photon APD operation is described, including the occurrence of micro plasma in avalanche breakdown. Micro plasma sites are bi stable below their saturation currents, and this is the regime where APD can be used. Above the saturation current level, however, micro plasma is self-sustaining and photoelectron multiplication is reduced, but the micro plasmas emit light.