This article reviews that advances in manikin software have enabled engineers to consider a fuller spectrum of user interactions with virtual products. It has been 15 years since Sammie—a computer model of a human or manikin—emerged from the research labs of Nottingham University in the United Kingdom, promising human-factors (HF) engineers a supporting software tool to improve the study of human elements of product design. Although Sammie incorporated accurate anthropometric data and representative joint constraints, the software was very difficult to use, could not import models from a computer-aided-design (CAD) system, and was not dynamic. After phase 1 of the collaborative project, Rolls-Royce and VSEL expanded their study to evaluate the use of virtual reality for the design and layout of larger and more complex machinery spaces. This second collaborative effort had several objectives: to understand how virtual-prototyping technology could help designers better visualize complex designs, design for ease of assembly and maintenance much earlier, train maintenance engineers, and enhance communications between disparate project teams, customers, and suppliers.
It has Been 15 years since "Sammie"-a computer model of a human, or manikin-emerged from the research labs of Nottingham University in the United Kingdom, promising human-factors (HF) engineers a supporting software tool to improve the study of human elements of product design. Although Sammie incorporated accurate anthropometric data and representative joint constraints, the software was very difficult to use, could not import models from a computer-aided-design (CAD) system, and was not dynamic.
Still, Sammie was a vast improvement for HF engineers, who had typically interpreted anthropometric tables themselves. The tables, which consisted of measurements of standard anthropometric variables derived from representative samples of human populations, defined anthropometry in terms of percentile information. (For example, a 95th-percentile man would be taller, stronger, or more intelligent, say, than all but 5 percent of the population.) With this information in hand, HF engineers would draw reach and visual envelopes or even construct two-dimensional cardboard manikins to replicate the seated, standing, and crawling movements of product users.
Subsequent advances in hardware and graphics technology have spawned three-dimensional manikins that can "inhabit" virtual environments created with the aid of a computer. These environments contain a digital representation of a prospective product that can then be tested for ease of assembly, operation, and maintenance.
The vast improvement in manikin software functionality has provided three essential benefits. With such software, the HF engineer can obtain easily digestible quantitative results of HF analyses. Moreover, the software reduces, and in some cases eliminates, the need for full-scale physical mock-ups. Finally, such software facilitates the communication of analysis data to all relevant stakeholders in a product-development project.
However, in the quest to develop the most advanced virtual manikin, many software developers have neglected the importance of the simulation environment itself. As a result, their human-modeling products address only half the problem. After all, product operation, equipment maintenance, and machinery access occur in complete environments, not simply in the subsets of an environment specially created for traditional HF analyses.
New approaches that attempt to place manikins in a much larger environment-one that is more typical of what an actual user of a product or system would encounter-are emerging, enabling manufacturers and customers to create an unambiguous vision of a product's appearance, operation, and maintenance.
The human being is often regarded as one of the most unpredictable components of any system and a major, usually limiting, design parameter. The environments and products that engineers design are inhabited, used, and maintained by human beings. And these users, of course, come in all shapes and sizes, with wide variances in strength and flexibility.
The importance of investigating the human element, driven increasingly by contractual mandate, has pushed manufacturers to consider user issues-from the point of view of the person who operates or maintains a product, for example, or who trains others in its use or maintenance-much earlier in the design process. In many application areas, there is a strong requirement to model the physical elements of the human alongside models of workplaces and equipment. Unfortunately, many of the human-modeling systems currently available are stand-alone applications that constrain the user to work with small assemblies, or partial representations of the target environment.
Moreover, many products are not integral components of the design environment. They are frequently used as stand-alone applications that require users to go off-line to perform human-centered analysis, creating a need for additional design time and expense. By contrast, DIVISION Ltd., with worldwide headquarters in Bristol, U.K., and U.S. headquarters in San Mateo, Calif., has developed its dV /Manikin module as an integrated component of its suite of interactive product simulation (IPS) software tools to facilitate design-orienting as well as design-verifying human-centered analysis. (For more information on interactive product simulation, see "The Virtues of Virtual Products," June.) The result is an integrated solution across an enterprise-wide design environment.
The development of dV/Manikin reflects significant changes in the way human-factors analyses are used by a wide spectrum of stakeholders both within a manufacturing enterprise and among its customer base. One of the most important changes is the migration of basic, or first-cut, human-factors analyses to the engineering desktop. Design engineers who want to ensure that their designs conform to approximations of human capability require sophisticated dV/Manikin human-modeling capabilities, based on a predetermined set of manikins (anthropometry for 5, 50, and 95 percentiles for both males and females) for posturing the manikin and generating reach envelopes and vision cones.
On the other hand, design specialists such as HF engineers generally require a full suite of dV/Manikin analyses (comfort, posture, biomechanics, center of gravity, reach, etc.) to undertake detailed, intensive human-factors analyses. Such analyses use manikins of any percentile definition, based on any anthropometric variables derived from either published or user-definable anthropometric databases.
Groups such as managers, customers, and suppliers require a human-modeling system capable of visualizing the result of human-centered analyses. This reduces the ambiguity associated with communicating the relationship between human and product throughout the enterprise. The HF group can forward the detailed human-centered analysis results to the relative stakeholders in the enterprise using the modular IPS environment.
This enables communication of human-centered design data vertically within the enterprise, and horizontally to external stakeholders (such as project managers, customers, and contractors).
The ability to test a design in terms of human issues (such as functional reach, vision obstruction, and height and access limitations) within a digital environment before a physical prototype is built provides a significant return on investment for many organizations, as the following examples suggest.
Virtually every major manufacturer in the aerospace, automotive, and heavy-engineering industries is using some form of human-modeling capability in its product design and development. One example is the work that Rolls-Royce has recently undertaken to make manikins an integral part of the design environment used to prove product maintainability.
At the end of 1997, Rolls-Royce completed a five-year comprehensive evaluation focusing on the use of virtual-prototyping technology. The previous three years had been spent in collaboration with Vickers Shipbuilding and Engineering Ltd. (VSEL) as part of the Virtual Reality and Simulation Initiative (VRS). VRS brought together a number of companies, including, besides Rolls Royce and VSEL, British Nuclear Fuel Ltd. (BNFL) and Fluor Daniel, to assess virtual reality (VR) in engineering and conm1ercial applications. Each participant had a particular objective. Rolls-Royce was investigating the technology to provide HF support and evaluate the ergonomic issues associated with Trent 800 maintenance. VSEL wanted to use the technology to support the modeling of nuclear submarines, and BNFL wanted a combined IPS and human-modeling environment to test control room designs.
An important objective of this early IPS/human-modeling study was to assess the potential of using virtual-prototyping technology to complement Rolls-Royce's CAD efforts. The engineering group believed that virtual prototyping was a logical stage between digital pre-assembly and the costly fabrication of a full-scale engine mock-up. A virtual engine, created in DIVISION'S dVISE software (used to create a virtual product and its environment) using CADDS5 assembly data, was manipulated by dV/ Manikin and tested for accessibility and disassembly/reassembly routines of pipes and pumps. The pipe routes were col or-coded and segmented into groups of objects, thereby allowing their removal and manipulation by engineers via the manikin. The final phase was to use the manikin as the manipulation medium.
The manikin provided visual feedback whenever acceptable limits of the normal range of movements had been exceeded. A series of on-screen indicator bars illustrated movements so that designers could understand the development of tasks in a certain area. Alternately, HF engineers could trace the movements of limbs or levels of vision through reach or sight envelopes. Due to the fidelity with which human motions and tasks were duplicated, the engineering team was able to prototype the Trent 800 maintainability procedures without the need for a full-scale engine mock-up.
Simulating a user's ability to see, reach, and maneuver is an integral part of the design process.
Visualizing Complex Designs
After phase 1 of the collaborative project, Rolls-Royce and VSEL expanded their study to evaluate the use of VR for the design and layout of larger and more complex machinery spaces. This second collaborative effort had several objectives: to understand how virtual-prototyping technology could help designers better visualize complex designs, design for ease of assembly and maintenance much earlier, train maintenance engineers, and enhance communications between disparate project teams, customers, and suppliers. Although the project is now completed, no results have been publicly released.
However, both Rolls-Royce and VSEL invested significantly in dVISE and dV/Manikin, while Rolls-Royce Aero Engines is still evaluating the results.
Rolls-Royce has extended its VR work to include the Westinghouse/R-R WR 21 gas turbine project, a collaboration with Northrop Grumman. WR 21 is the U.S. Navy's only propulsion development program for ships, and it is the base engine of choice for the twenty-first century, since it is the U.K. contender for the Common New Generation Frigate.
"The human-factors design challenges with all our marine power products are the usual problems of having large pieces of equipment in small spaces," said Tom Mayfield, human-factors consultant with Rolls Royce and Associates in Derby, U.K. "Wherever you have fairly high-tech equipment requiring some level of human intervention, the ability to see, reach, and carry out the tasks using tools is an integral part of the design process." The WR 21 module, 15 feet high and 16 feet long, has been developed with low maintenance costs as a prime focus, resulting in a modular engine.
The use of manikins produces reach and visual envelopes that indicate the extent to which a maintenance engineer would be able to gain access to the key elements of the equipment and actually manipulate the necessary tools and objects. So far, the results are promising, according to Mayfield. A similar project, now going into its design phase, will integrate manikins into Rolls-Royce's development of the nuclear steam-raising plant for a new class of British nuclear submarines.
As the components that make up a nuclear power plant are ultimately assembled in the confined space of the hull, this stage, along with other regular activities such as refueling and defueling, are prime HF applications for manikins. Mayfield has incorporated dV/Manikin in situations in which the manual handling of equipment is required, such as the manipulation of long-handled tools for opening and closing hatches and valves within acceptable limb and joint radii. The manikin generates a wide variety of reach and visual arcs.
"Virtual-world manikins, such as dV/Manikin, provide a relatively accurate range of sizes that you would find in the population," Mayfield said. "This provides readily understandable feedback for designers, who have typically designed for the nonexistent average man."
This extension of fundamental HF work beyond the traditional core of specialists is due to improvements in the usability of human-modeling software, combined with the increased importance of considering the human early on in the design process. Continued adoption of human-modeling technology will likely be based on the need to perform human-centered analysis within the context of the complete virtual product. Both the WR-21 and nuclear-propulsion projects require the manikin to be positioned within very large amounts of CAD geometry, and manipulated in real time—requirements that have heretofore inhibited the widespread deployment of human-mode ling technology. VR producers are concerned with improving the quality of their manikins and with attempts at incorporating existing anthroprometric packages into their software.