More Like this

1. Analysis of Perceived Social Impacts of Existing Products Designed for the Developing World, With Implications for New Product Development

   https://doi.org/10.1115/1.4044323  

   Ottosson, Hans J.; Mattson, Christopher A.; Dahlin, Eric C. (May 2020, Journal of Mechanical Design)

   Abstract Engineered products often have more social impacts than are realized. A product review was conducted to bring this to light. In this paper, we show the extent to which different social impacts in 11 impact categories are co-present in 150 products and how this can help engineers and others during the product development process. Specifically, we show how social impact categories not previously considered can be identified. The product review resulted in 13,200 data points that were divided into two data sets, one with 8800 data points from which a social impact probability table was created. The remaining data points were then used to validate the table. All data points were then combined to create a final social impact probability table. This table provides insight for how various social impact categories correlate and can assist engineers in expanding their views to include additional social impact objectives and thus achieve a design with broader social impact or a design with minimized unwanted negative social impact. A simple method for predicting social impact is also created in order to assist engineers when developing products with social impacts in mind. 

   more » « less
2. A Graph Neural Network Approach for Product Relationship Prediction

   https://doi.org/10.1115/DETC2021-69462  

   Ahmed, Faez; Cui, Yaxin; Fu, Yan; Chen, Wei (August 2021, ASME 2021 International Design Engineering Technical Conferences & Computers and Information in Engineering Conference)

   Abstract Graph representation learning has revolutionized many artificial intelligence and machine learning tasks in recent years, ranging from combinatorial optimization, drug discovery, recommendation systems, image classification, social network analysis to natural language understanding. This paper shows their efficacy in modeling relationships between products and making predictions for unseen product networks. By representing products as nodes and their relationships as edges of a graph, we show how an inductive graph neural network approach, named GraphSAGE, can efficiently learn continuous representations for nodes and edges. These representations also capture product feature information such as price, brand, and engineering attributes. They are combined with a classification model for predicting the existence of a relationship between any two products. Using a case study of the Chinese car market, we find that our method yields double the F-1 score compared to an Exponential Random Graph Model-based method for predicting the co-consideration relationship between cars. While a vanilla Graph-SAGE requires a partial network to make predictions, we augment it with an ‘adjacency prediction model’ to circumvent this limitation. This enables us to predict product relationships when no neighborhood information is known. Finally, we demonstrate how a permutation-based interpretability analysis can provide insights on how design attributes impact the predictions of relationships between products. Overall, this work provides a systematic method to predict the relationships between products in a complex engineering system. 

   more » « less
3. Social Impact in Product Design, An Exploration of Current Industry Practices

   https://doi.org/10.1115/1.4045448  

   Pack, Andrew T.; Rose Phipps, Emma; Mattson, Christopher A.; Dahlin, Eric C. (July 2020, Journal of Mechanical Design)

   Abstract Though academic research for identifying and considering the social impact of products is emerging, additional insights can be gained from engineers who design products every day. This paper explores current practices in industries used by design engineers to consider the social impact of products. Forty-six individuals from 34 different companies were interviewed to discover what disconnects exist between academia and industry when considering a product’s social impact. These interviews were also used to discover how social impact might be considered in a design setting moving forward. This is not a study to find “the state of the art,” but considers the average engineering professional’s work to design products in various industries. Social impact assessments (SIA) and social life cycle assessments (SLCA) are two of the most common processes discussed in the literature to evaluate social impact, both generally and in products. Interestingly, these processes did not arise in any discussion in interviews, despite respondents affirming that they do consider social impact in the product design. Processes used to predict social impact, rather than simply evaluate, were discussed by the respondents. These tended to be developed within the company and often related to industry imposed government regulations. To build on this study, the findings herein should be further validated for executives, managers, and engineers. A study specific to these roles should be designed to understand the disconnect better. Additionally, processes should be developed to assist engineers in considering the social impact of their products. Work should also be done to help educate engineers and their leaders on the value of considering the social impact in product design. 

   more » « less
4. Combining Direct and Indirect User Data for Calculating Social Impact Indicators of Products in Developing Countries

   https://doi.org/10.1115/1.4047433  

   Stringham, Bryan J.; Smith, Daniel O.; Mattson, Christopher A.; Dahlin, Eric C. (December 2020, Journal of Mechanical Design)

   Abstract Evaluating the social impacts of engineered products is critical to ensuring that products are having their intended positive impacts and learning how to improve product designs for a more positive social impact. Quantitative evaluation of product social impacts is made possible through the use of social impact indicators, which combine the user data in a meaningful way to give insight into the current social condition of an individual or population. Most existing methods for collecting these user data for social impact indicators require direct human interaction with users of a product (e.g., interviews, surveys, and observational studies). These interactions produce high-fidelity data that help indicate the product impact but only at a single snapshot in time and are typically infrequently collected due to the large human resources and cost associated with obtaining them. In this article, a framework is proposed that outlines how low-fidelity data often obtainable using remote sensors, satellites, or digital technology can be collected and correlated with high-fidelity, infrequently collected data to enable continuous, remote monitoring of engineered products via the user data. These user data are critical to determining current social impact indicators that can be used in a posteriori social impact evaluation. We illustrate an application of this framework by demonstrating how it can be used to collect data for calculating several social impact indicators related to water hand pumps in Uganda. Key to this example is the use of a deep learning model to correlate user type (man, woman, or child statured) with the raw hand pump data obtained via an integrated motion unit sensor for 1200 hand pump users. 

   more » « less
5. A Method for Creating Product Social Impact Models of Engineered Products

   https://doi.org/10.1115/1.4044161  

   Stevenson, Phillip D.; Mattson, Christopher A.; Dahlin, Eric C. (April 2020, Journal of Mechanical Design)

   Abstract All products impact the lives of their users, this is called social impact. Some social impacts are commonly recognized by the engineering community, such as impacts to a user’s health and safety, while other social impacts can be more difficult to recognize, such as impacts on families and gender roles. When engineers make design decisions, without considering social impacts, they can unknowingly cause negative social impacts. Even harming the user and/or society. Despite its challenges, measuring a program’s or policy’s social impact is a common practice in the field of social sciences. These measurements are made using social impact indicators, which are simply the things observed to verify that true progress is being made. While there are clear benefits to predicting the social impact of an engineered product, it is unclear how engineers should select indicators and build predictive social impact models that are functions of engineering parameters and decisions. This paper introduces a method for selecting social impact indicators and creating predictive social impact models that can help engineers predict and improve the social impact of their products. As a first step in the method, an engineer identifies the product’s users, objectives, and requirements. Then, the social impact categories that are related to the product are determined. From each of these categories, the engineer selects several social impact indicators. Finally, models are created for each indicator to predict how a product’s parameters will change these indicators. The impact categories and indicators can be translated into product requirements and performance measures that can be used in product development processes. This method is used to predict the social impact of the proposed, expanded U.S. Mexico border wall. 

   more » « less