The Role of Virtual Ergonomic Simulation to Develop Innovative Human Centered Products

Abstract

The paper concerns the use of integrated methodologies and tools to perform innovative human centered development of products. Digital simulation of ergonomics by means of DHM is shown together with advanced tools for design, taking into account Knowledge-based systems, Design Automation and design of highly customized goods. Two different applications of the proposed approach are described, the first refers to an industrial product, the second to the medical domain. Both applications, even if belonging to completely different fields benefit from putting the human at the center of the developing paradigm from the very first step of product development. Some results and discussion highlight benefits and limitation of the approach and of the adopted tools.

1 Introduction

Digital Human Modeling (DHM) enables designers to face several design issues, such as comfort and posture prediction [1] for envisaged groups of users, task evaluation and safety, visibility of products [2] (machinery, equipment, vehicles, etc.), reaching and grasping of devices (buttons, shelves, goods, etc.) [3], and multi-person interaction to analyze if and how multiple users can cooperate among them and with the product. DHM can be used for product ergonomics [4], crash testing, product virtual testing, workplace design, and maintenance allowing a fast redesign and reducing the need and realization of costly physical prototypes, especially for those applications dealing with hazardous or inaccessible environments.

The aim of this work is to present a method and different ways in which DHM can be used to improve product development process, obtaining either flexible design tools or highly customized products. The paper, after a presentation of the method shows two applications in different domain to clarify the potentiality of DHM integrated with other design methods and tools.

2 Methodology

The potential of digital humans at a methodology level can be exploited to define a new design framework for managing new product development processes and product innovation (Fig. 1).

Fig. 1.

Human centered methodology framework

Actually, the capability to handle human factors in a design activity constitutes a deep change crossing many existing methods and tools.

Human centered design can be the starting point for the conceptual design phase. Technicians can start dealing with human data from the very first step of product development to embed customer’s personal requirements at the beginning of the pipe rather than at the end. This is the base also for a broad Computer Aided Ergonomic approach [5] to design the environment in which operators will carry out their tasks, no matter which is the application sector (e.g., industry, medical, transportation) who is the operator (taking into account also his/her skills, characteristics and culture) and what is the task to be performed. Tools available on the market, such as Siemens Jack [6] or Dassault System Delmia [7], are perfectly suited for most of the ergonomics issues and allow users to addresses practically any task of an industry-oriented scenario. By the way, there are still some open challenges for a better characterization of human models in terms physical details and of their behavior.

By embedding human factors in product design, a deep mass customization can be achieved. To be competitive on the market while delivering one to one personalized goods and services, the underlying methods and tools used must meet high efficiency standards though being adaptable, and so must be the production systems. Modularity, scalability and flexibility are the main drivers to be applied both to products and to production systems to gather a higher level of customization. For instance, design tools would benefit from an improved function-oriented modularization together with a fluid data exchange though the entire product development process.

Another key issue strictly connected to enhanced customization is the way in which design tools can be integrated and used automatically within a knowledge-based framework. Actually, time and costs can be dramatically reduced by changing design paradigm according to Design Automation (DA) prescriptions. The main features of a good DA practices are Advanced Modeling, Parametric/Constraint Management and Integration of Iterative Steps/Interactive Analysis. Moreover, Knowledge Based systems can be exploited to embed senior technicians’ experience into an automatic design environment using, for instance, Finite Element simulation tools [8]. The capability of a company to quickly define the most suitable strategy and to implement a light and effective product development process paradigm is the key to success.

3 Applications

The paper shows two different applications in the medical and industrial fields in which many of the cited methods have been integrated. The applications refer to the ergonomic-oriented design of supermarket display units and to the custom-fit design of artificial lower limbs.

The industrial case study is a refrigerated display unit of the type used in supermarkets and groceries. The display unit design is generally focused to meet goals and constraints centered on product, such as storage capacity or product visibility. What is missing, or at least underestimate, is the role of people in the buying process. Therefore, in a human centered approach, the display unit will have to meet as much as possible the requirements of a variegated population of users characterized by different anthropometric measures. Moreover, besides customers picking up goods during opening time, also maintenance technicians and workers filling out the shelf with fresh products must be considered. For each category of people interacting with the display unit, some ergonomic aspects are more relevant than others: i.e., visibility and reachability of goods for customers, reachability of some display components for technicians and, the most important, posture and stress for operators who repeat the same task for hours and may suffer from musculoskeletal disorders. An integrated method based on parametric modeling and DHM simulation of human-machine interaction provides the designers with a tool that updates ergonomic evaluation indexes at each modification of the display unit design.

The medical case study refers to the design and test of lower limb prosthesis required by a person who had a leg amputated. The artificial leg is composed by standard parts (e.g., knee, foot, connectors) and by the socket that is a highly customized part. In this case, the patient is already the main reference for any design step, but the weakest link in the design process is the lack of digitalization of the available tools. Actually, in most of orthopedic labs the realization of sockets is still accomplished in a manual way. What is proposed according to the novel approach is to switch to a complete virtual approach and automated design, while taking into account that the final users will be an orthopedic technician who is not supposed to have a deep expertise in IC technologies. Input data are generated at the beginning of prosthesis development by acquiring patient’s geometry with a MRI scanner. After that, the design module has the goal to (i) create the virtual socket relying on patient’s anatomy and (ii) define a consistent assembly of standard components chosen accordingly to patient’s history. The test module has the goal to forecast the prosthesis behavior when used by the patient and different kind of numerical simulation (FE analysis and gait simulation with DHM) are performed almost automatically to check comfort and performance of the final product. All the knowledge related to routine steps of both design and test have been embedded into the system and the human intervention is required only to control the procedure and evaluate partial and final outputs.

The two different applications are better described and discussed in the followings.

3.1 Industrial Application

This application refers to a vertical display unit and to its ergonomic evaluation method based on a full virtual approach. We propose a method based on DHM commercial tool and a CAD software in order to evaluate in details accessibility and visibility of products, generating results that are both numerical and graphical for an easy comprehension and usability.

Human modeling tools are used to simulate the person physically interacting with the product and the environment. This allows a quantitative assessment of accessibility and visibility of products, as requested in this specific application, but, for instance, posture analysis and tasks evaluation can be carried out with the same approach.

The application exploits the potential of a parametric approach for both modelling of the product people is going to interact with, and the digital human models. Depending on the case of use, the parameters to be considered can be different: the product may be designed so that any potential geometric variant is achieved only modifying a few independent variables; human models can be varied as well, for instance, depending on gender, age, height and build. In this case, we adopted two software solutions by Siemens: Solid Edge ST4 for the CAD side and Jack as DMH solution.

When dealing with a display unit it is very important to have a flexible approach. On one side, the supermarket is visited by every kind of people and the larger portion of population is satisfied by the buying experience, the better it is. On the other side, the display unit can be easily configured by repositioning shelves at different heights or using shelves with a different depth. These variations impact on the way people will be able to see and to reach product they want to buy, and thus the efficiency of the overall process.

Even if designers know very well generic issues, or can easily forecast some of them, (e.g., the highest shelf is the less reachable) it is not easy at all to gather a quantitative index measuring the comprehensive change of performance of the product/human system whenever one or more parameters are changed. The following examples show how this can be done and how a graphical representation of results can improve usability of results during product development.

The example reported in Fig. 2 addresses the visibility of a vertical, open (without doors), display unit. Colors are used to highlight the hidden zones for respectively a man whose height correspond to the 95th percentile, and a woman corresponding to the 25th percentile of the European population. By changing avatars’ height, distance from the display unit and shelves vertical position it is possible to quickly gather a new configuration and the related visibility results.

Fig. 2.

Open display unit visibility test for different human models (Color figure online)

When dealing with reachability the test campaign can be much more complex respect to visibility because it imply a study on the posture assumed while measuring the maximum reachable distance. Moreover, different postures can lead to different results in terms of both distance reached and comfort [9].

Another notable example of exploiting parametric models to gather ergonomic results consist in assessing the reachability of the horizontal surface of a shelf in a display unit with doors. Actually, for any point on a 50 × 50 mm grid in which each shelf has been divided, the avatar assumes a different position and it may interfere with the display. Figure 3 shows the results of a reachability analysis in which collision (red cells) or proximity (yellow cells) are detected among different parts of the display unit and of the avatar. Mapping the results using colors directly on the 3D model of the product allows technicians to easily and quickly interpret simulation outcomes. The results can be obtained for each shelf and varying the digital human model to cover the desired target of population.

Fig. 3.

Closed display unit reachability test and collision detection. The cells colors mean: Gray: not reachable; Red: collision; Yellow: proximity; Green: good reachability. Letters inside the yellow and red cells refer to the colliding parts: H: head; A: arm; F: frame; S: shelf; D: door (Color figure online).

3.2 Medical Application

This example of methods and tools integration refers to the design and simulations of lower limb prosthesis, with particular attention to the most customized component i.e., the interface between the patient’s residual limb and the artificial leg, namely, the socket.

In this context, the main goal has been to implement an automatic design and simulation procedure. The rationale beside this goal is, uncommonly, not just to save time in performing the computer-based design process, but to let it be feasible. Orthopedic technicians and prosthetists, actually, have not got the competences required to handle CAD systems and simulation tools required to virtualize the process of making a socket. Thus, the need for a system that highly support the technicians to generate the geometry of the socket and automatically runs FE analysis to validate it, is crucial.

In order to obtain a proper design platform the point of view is shifted from the tools available to the final user of the socket, or, better, to its 3D digital avatar. A huge research effort allowed to gather an environment in which a non-technical user is able, exploiting his/her medical expertise, to obtain the design of the socket, the results of the simulation of socket-residuum interaction, and the chance to apply the entire prosthesis to a digital avatar to analyze the gait.

The design platform integrates tools to acquire patient’s data, CAD tools to model prosthesis components, a finite element analysis (FEA) package to study the socket-residual limb interaction, and a digital human modeling (DHM) system to perform gait analysis. It consists of two main environments (Fig. 4): (i) the prosthesis modeling laboratory (PML) and (ii) the virtual testing laboratory (VTL). The PML allows the prosthetist to design the whole prosthesis, for both transtibial and transfemoral amputees. The 3D socket model is created onto the residual limb digital model using a dedicated CAD tool (Socket Modeling Assistant (SMA)) and the standard components are appropriately selected and modeled with a commercial 3D CAD software according to patient’s needs. The VTL interacts with PML to assess the prosthesis design before manufacturing. It permits to evaluate automatically or semi-automatically the socket shape thanks to the integration of numerical simulations tools. Furthermore, by the use of a DHM system, it allows to virtually set up the artificial leg and simulate patient’s posture and walking, validating prosthesis functionality and configuration.

Fig. 4.

Prosthesis design platform

Within the VTL, the numerical simulation has been implemented embedding a FEA solver. Among the various FE solvers commonly used in this field, we adopted Abaqus package and use it through the Abaqus Scripting Interface for executing a series of “jobs” without user’s intervention.

Once the prosthetist has created the 3D socket model, the system acquires the input for the analysis, produces the files required to generate the FE model and calls the module to execute the analysis. The FE model is automatically created; Abaqus solves the analysis, and generates the file output containing the pressure values. These are imported in SMA and visualized with a color map. SMA evaluates pressure distribution and highlights the areas that should be modified. Figure 5 shows an example of the geometric modelling of the socket (Fig. 5a) and the pressure map identifying the load zones to be evaluated (Fig. 5b). Thus, technicians are only requested to evaluate the level of intensity of the modifications to be applied to the geometric model before running the few simulation requested until the result is satisfying. Such a process, in the manual procedure, implied the production of at least one check socket and some trials with the patient, while in this new way the patient is involved only when the design of the socket is much more likely to be correct.

Fig. 5.

Socket model (a), Socket/residuum pressure map obtained by FE analysis (b)

Moreover, having the digital models of all the components of the prosthesis allows introducing new manufacturing technologies, as the Fused Deposition Modelling 3D printing we are testing to create a fast and reliable product while preserving full customization for each patient.

Another important aspect of the change of paradigm in leg prosthesis design is related to the use of DHM for the prosthetized patient. Human modelling tools, actually, could be used to analyze the virtual gait of the patient’s avatar wearing his/her virtual leg before it is manufactured. Figure 6 shows an image of the model of the patient in which the left leg has been replaced by the prosthesis.

Fig. 6.

Model of the patient having the residuum of the limb reconstructed by RMI Dicom images (a), assembled with the custom socket and standard components (b) to create the complete model (c) to be used for gait analysis (d).

The analysis of the gait of the patient can be performed either on the virtual avatar or on a patient already having and using a lower limb prosthesis. In this last case, the Motion Capture (MoCap) techniques can be used to gather gait data. Optical markerless low cost sensors have been tested for this aim and results can be found in [10].

Since a patient walking with a prosthesis may encounter a set of well-known and classified abnormalities respect to a standardized correct gait, we defined a module to support expert personnel in identifying these abnormalities so that the prosthesis can be tuned at its best. To this aim two sources of knowledge have been used. First, we formalized the knowledge of the Atlas of Limb Prosthetics [11] and embed it in the module to automatically assess key parameters of the gait, compare those values with a set of reference values and highlight eventual deviations for each specific case. The second set of knowledge was obtained directly from orthopedic technicians who are used to correct design issues and to fine-tune the regulations of the prosthesis in order to empirically find the optimum configuration to ensure comfort and safety. The developed module has the goal to support and ease decisions by means of automatic quantitative evaluations, it is not aimed at substituting the work of the prosthetist. Actually, s/he must take into account also other psychological or non-technical issues by creating trust and an intimate link with the patient that will never be fully replaced by an automated procedure.

4 Discussion and Conclusions

The results gathered with this work are still at a methodological level and the case study have the goal to clarify potentiality of the integrated approach but they do not represent a validation. By the way, it is undeniable that shifting the focus on human models each time a person interact with a machine brings notable benefits. Ergonomics can be brought inside the product development process from the conceptual design so that main issues can be prevented. The exploitation of mature technologies such as parametric CAD together with DHM may bring to huge improvement in time saving and, thus, permit deeper analysis. The first case presented deals with this kind of applications and shows a way designers may adopt to shorten the test campaign that, otherwise could be too much time consuming.

The second application, on the other hand, is not addresses at refining the use of existing tools in an industrial context, but it refers to a medical domain where the goals are completely different. The case shows the complex work of designing and testing lower limb prosthesis with the aim of highlighting tools integration (either commercial or specifically developed) while constructing a new design paradigm. Actually, in this application the result consists in being able to substitute the traditional manual process with a full virtual approach. To this aim, DHM was used together with Design Automation prescriptions to simulate the behavior of the socket component. A knowledge-based approach has been adopted to embed orthopedic notions and technicians’ experience into a system able to detect gait abnormalities and suggesting how to fix them. Using digital data instead of physical copy of patient’s body segments allows storing human data and creating a database, for instance, of residual limb geometries and gait motion characteristics, so that they can be reused or set as a reference. The digital representation opens the way also to flexible manufacturing technologies, such as 3D printing, that can cut costs while preserving the required level of customization.

Several other potential integrations among the methods shown in Fig. 1 are feasible and may vary depending on the specific aim and domain. Even if some barriers exist (data exchange, personnel training, and psychological resistance) and are slowing down this process, in the future many other works will exploit the centrality of DHM and ergonomics to improve product development process.