For millennia the horse was the primary mode of transportation for mounted soldiers. Ingress and egress from a horse’s back is straightforward, space claims are only related to the size of the saddle, and there were no confining walls to restrict what soldiers carried while on horseback. With the rise of the modern mechanized army, vehicle design became more complex. Critical to the effective design of vehicle interiors is an accurate model of the encumbered operator or passenger. Developments in three-dimensional (3d) scanning, computer-aided design (CAD) and other model creation capabilities make it possible to reproduce accurately the underlying human form and to add equipment encumbrances. This paper relates approaches taken in studies where Soldiers or aviators were modeled to define space requirements or reaches. Details of the modeling process, validation, and study results are given. Future research is discussed. INTRODUCTION Archeological evidence suggests the horse was first domesticated approximately 4000-3500 BCE in the Eurasian Steppes [1] [2] [3]. The added power, size, and mobility a horse provided was quickly adapted to gain an advantage in conflicts [4]. Whether ridden or pulling a load, horses remained the mainstay of mounted armies until the 20th century. It is relatively easy to get on and off a horse. Invention of stirrups and saddle helped in that regard. The limit of what could be taken on horseback was determined by the size and shape of the rider and what could be strapped to the horse itself [5]. Thus, ingress, egress, and defining load space were straightforward. That all changed when horses were replaced by motorized vehicles. Walls or other structures confined occupants and constrained what may be either brought inside a vehicle as freestanding equipment or worn by an occupant. Door and hatch size became critical for efficient ingress and egress. Thus, vehicle interior design today is a complex trade-off between competing priorities. A useful heuristic for partitioning the problem space of vehicle interior design is to consider anatomy, geometry, and physics of the system. Anatomy refers to systems worn for protection from threats (blast overpressure, ballistic, translational, thermal, chemical, etc.). Geometry considers volume in terms of interior volume (occupant space, workstation space, reaches, etc.) and the volume the human occupies. Physics refers to dynamics, that is volume and its mass in motion. In this paper we are interested in analysis of volume and mass. Proceedings of the 2009 Ground Vehicle Systems Engineering and Technology Symposium (GVSETS) Approved for public release; distribution is unlimited. No Hatches on Horses: Creating Accurate Encumbered Digital Human Models To Assess Operator Or Passenger Space Requirements Page 2 of 7 Representation of the human form has improved with developments in three-dimensional (3d) surface scanning and other digitizing modalities. Powerful computer graphics allow complex environments, like vehicle interiors, to be created and visualized digitally. These modeling and simulation capabilities make it possible to evaluate and weigh the impact of design decisions. CREATING APPROPRIATELY SCALED DIGITAL HUMAN MODELS Development and validation of digital human models (DHM) is the key component to effective vehicle interior modeling and simulation. The major commercial digital human modeling software packages used for human factors evaluations such as Jack (Siemens PLM), RAMSIS (Human Solutions), Delmia-Human (formerly Safework), as well as newcomers such as Santos (University of Iowa/ESI) provide premade DHMs whose size/shape may be selected based on percentile score or other criteria. They also allow creation of new forms from anthropometry generated by outside analysis. Selection of body dimensions to drive DHM size/shape must be made with care. Pre-existing datasets and percentile ranges may or may not cover the population in question. When user-generated anthropometry input is available, it is best to enter as many body dimensions as possible to minimize the number of body dimensions estimated from built-in regressions. Regressions, like the percentile values, may have been computed from a population different from the target population of study. For example, the Delmia V5 Human takes up to 103 measurements and imputes missing values based on 1988 US Army Anthropometric Survey (ANSUR) regressions [6]. Further, given that most evaluations utilize models at the extremes of accommodation, securing the appropriate representative anthropometric data for a user population becomes paramount. The phrase “5th to 95th percentile” has entered the engineering lexicon as shorthand to describe a 90% population accommodation envelope. A percentile is a univariate value; however, seat adjustments, displays, controls, and the like must accommodate many body dimensions simultaneously to be effective. As [7] and others have pointed out, when simultaneous accommodation is required, univariate percentiles are not appropriate to set boundary conditions [8] [9]. It is better to establish body size/shape parameters with multivariate statistical analyses such as Principal Components Analysis (PCA) or Factor Analysis (FA). A number of papers are available that demonstrate the application of PCA to establish worse-case anthropometry for engineering applications [10] [11] [12]. Generating a custom DHM for a given application was outlined by [7] for the Jack V4 human figure. The approach is not unique to the Jack system and may be applied to DHMs in other simulation environments with minor changes to accommodate input variables. Essentially, the method is to input target anthropometry into the simulation’s data editing feature and adjust the resultant DHM if necessary (Fig. 1). The adjustment step is critical. Some of the dimensions are cumulative. For example, in Jack sitting height is influenced by several variablessitting eye height, acromial sitting height, and elbow height. If one of the constituent values is changed, the program may change the other values according to a built in regression function. Segment values may be locked but some small change may be required to get all segments to link up correctly. The markers are stored with the DHM as a visual reference check against unintentional changes in dimensions (Fig. 2). Once all anthropometric data have been input a final check should be done to ensure all segment lengths are correct. Another approach to DHM creation is to use a threedimensional (3d) whole body surface scan as a template against which the DHM is scaled. Anthropometry may be extracted from the scan or measured directly and input as described above. The resulting DHM is compared to the scan to check how closely they match (Fig. 3). Adjustments are made to the DHM until the desired level of fit is achieved. Figure 1: Jack V4 anthropometry data input sheet. Proceedings of the 2009 Ground Vehicle Systems Engineering and Technology Symposium (GVSETS) Approved for public release; distribution is unlimited. No Hatches on Horses: Creating Accurate Encumbered Digital Human Models To Assess Operator Or Passenger Space Requirements Page 3 of 7 CREATING ENCUMBERED DIGITAL HUMAN MODELS While clothing and encumbrances are disregarded for many types of office and commercial workspace designs, they are an important factor to consider for military systems where space is often at a premium and the additional clothing and equipment can add significant weight and bulk to each individual. Some typical examples include multilayered ensembles that provide protection against nuclear, biological and chemical threats, clothing to operate in extreme cold weather environments and body armor for ballistic and fragmentation protection. Additionally, load bearing vests and packs are worn to help transport sustainment supplies, along with advanced tactical equipment such as communication gear, components for night-vision and thermal imaging capability, as well as lasers for range-finding and target designation. Clearly, clothing and equipment items of this nature must be accounted for in the workspace designs of military systems if specified accommodation goals are to be met. Creating an encumbered DHM has been a challenge in the past. Often, 3d digital equipment models were not available. If models were available, getting them into the correct format, the appropriate resolution and scale, and attaching them to the DHM was not easy. Recently, many of the obstacles have been overcome. A concerted effort is being made to create 3d models of clothing & equipment commonly worn by Soldiers. A drag-and-drop capability to place items in an approximate position is available for most DHM environments. What are the items and where should they go? To answer the questions we have turned to 3d whole body surface scanning. Defining load components and how they are distributed on the body is a task for a subject matter expert (SME). Figure 4 illustrates SME defined equipment and its distribution on a Soldier body relative to the digital model. Photographs are one means to capture where items are worn on the body, but a direct comparison between an encumbered Soldier and his/her simulated counterpart is better. Such a comparison is possible if a 3d whole body surface scan of the encumbered subject is set as a reference. As was the case when a semi-nude scan was used to define DHM size/shape, a 3d surface scan and associated clothed anthropometry may be used to generate an encumbered model. The progression from unencumbered to encumbered DHM is illustrated in Figure 5 for a standing figure. The same approach may be used to create a seated DHM. Reference markers are used to maintain correct anthropometry. Table 1 provides results from a test case where a Jack DHM was matched to an encumbered 3d scan. Measurement error values from a study of clothed anthropometry [13] serve as a refere
[2]
Jeffrey A. Hudson,et al.
A Multivariate Anthropometric Method for Crew Station Design: Abridged
,
1993
.
[3]
Steven Paquette,et al.
The Effects of Multilayered Military Clothing Ensembles on Body Size: A Pilot Study
,
1999
.
[4]
Karen Young,et al.
Derivation of Boundary Manikins: A Principal Component Analysis
,
2008
.
[5]
J. Chamberlin.
Horse: How the Horse Has Shaped Civilizations
,
2006
.
[6]
Robin Bendrey,et al.
The Earliest Horse Harnessing and Milking
,
2009,
Science.
[7]
Gregory F. Zehner,et al.
Verification and Validation of Human Modeling Systems
,
2008
.
[8]
R. Reyment,et al.
Multivariate Morphometrics, 2nd Edition.
,
1985
.
[9]
William F. Moroney,et al.
Empirical Reduction in Potential User Population as the Result of Imposed Multivariate Anthropometric Limits.
,
1972
.
[10]
David W. Anthony,et al.
The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World
,
2008
.