Texas Instruments (TI) currently offers a comprehensive curriculum to supplement their Robotics System Learning Kit MAX (TI-RSLK MAX). While designed for all college students, TI would like to provide a modification of the curriculum specifically for first-year engineering design courses. The goal of this project was to modify the current TI-RSLX MAX curriculum, making it easier to digest for first-year engineering students and more applicable to shorter duration projects (∼8 weeks), while still being information dense. To better understand the user experience of the current TI-RSLK MAX curriculum and determine what users would want from a modified curriculum, user interviews were conducted and analyzed using inductive qualitative analysis techniques. Results showed that while the current TI robotics project has many positives, first-year students primarily struggled with the coding aspect of the project. These results directed and informed the prototype development of a modified curriculum module focused on teaching students code. Low- and medium-fidelity prototypes were developed and tested with users. Data from the usability study were analyzed using descriptive analysis techniques and results indicate that the medium-fidelity prototype helped students learn robotics coding. This research has implications for a high-fidelity modified TI-RSLX MAX curriculum and robotics education for first-year engineering design students.

How can instructors leverage assessment instruments in design, build, and test courses to simultaneously improve student outcomes and assess student learning to improve courses? A Take-away is one type of assessment method. It is unstructured text written by a student in AME4163: Principles of Engineering Design, the University of Oklahoma, Norman, US to record what they understand by reflecting on authentic, immersive experiences throughout the semester. The immersive experiences include lectures, assignments, reviews, building, testing, and a post-analysis for the design of an electro-mechanical system to address a given customer need. In the context of a Take-away, a student then writes a Learning Statement. The Learning Statement is a single sentence written as a triple, i.e., Experience|Learning|Value. Over the past three years at the University of Oklahoma (OU), we collected about 18,000 Take-aways and 18,000 Learning Statements from almost 400 students. In our earlier papers, we primarily concentrate on analyzing students’ Learning Statements by a text mining framework. In this paper, we focus on analyzing students’ Take-aways data using a Latent Dirichlet Allocation (LDA) algorithm, and then relate the Take-away data to the instructor’s expectations using text similarity. By connecting and comparing what students learned (embodied in Take-aways) and what instructors expected the students to learn (embodied in the course booklet), we provide evidence-based guidance to instructors on improving the delivery of AME4163: Principles of Engineering Design. The proposed method can be generalized to be used for the assessment of ABET Student Outcomes 2 and 7.

It has been well established that prototyping is an important part of development projects. Prototypes are used in engineering design courses but teaching students the strategic skills needed to decide what to prototype during development projects is often not a part of the curriculum. This paper proposes a model that uses the prototyping process to link individual prototyping activities and project goals. The proposition is that when designers work intentionally with prototyping as a process, they will be able to make strategic decisions about how and when to use prototypes. To teach students a prototyping process mindset, a simple prototype prioritization tool, the Prototyping Target, was developed and integrated into project-based product development courses. Fifteen student teams used the Prototyping Target to decide which project questions to answer with prototypes every week over three weeks. Results show that criticality to project success is a good starting point but not the only parameter to consider when deciding what to prototype. A comparison between prototyping plans made with and without the Prototyping Target shows that student teams selected a clearer purpose for their prototypes when they used the prioritization tool before planning specific prototyping activities. Overall, this paper demonstrates the benefits of teaching student designers to navigate the prototyping process.

Goal congruence, defined as agreement by all members of a team on a common set of objectives, has been positively associated with team cohesion, team performance and team outcomes, including grades earned. Yet there is little in-depth study at scale and across types of engineering design and innovation classes in higher education that examines the goals students set for their work together. This research explores goal congruence in 857 teams involving 1470 students across 18 classes over four years. To examine goal congruence, we use student assessments of their level of agreement on their goals as well as evaluations of their written goal statements. Machine learning techniques are used to automatically identify goal types and congruence between goals. We find that goal congruence on student teams is relatively low, even when they assess it as high, partly due to variety in the types of goals they identify. We categorize the goals students articulate for their teams into grade-, completion-, teaming-, learning-, problem-, output- and outcome-oriented goals and report variance in the types of goals identified in different pedagogical settings. Our findings have implications for how faculty design their classes, link learning outcomes to team projects and facilitate goal setting on student teams.

Empathy plays an important role in designers’ ability to relate to problems faced by others. Several researchers have studied empathy development in engineering design education; however, a majority of this work has focused on teaching designers to empathize with primary users. Little attention in empathy development research is given to empathizing with those affected in a secondary and tertiary capacity. Moreover, little research has investigated the role of students’ empathy in influencing their emphasis on sustainability, especially in the concept evaluation stage. Our aim in this paper is to explore this research gap through an experimental study with engineering students. Specifically, we introduced first-year engineering students at a large public university in the northeastern United States to a short workshop on sustainable design. We compared changes in their trait empathy and attitudes towards sustainability from before to after participating in the workshop. We also compared the relationship between students’ trait empathy, attitudes towards sustainability, and the self-perceived sustainability of their solutions in a design task. From our results, we see that students reported an increase in their beliefs and intentions towards sustainability and a decrease in their personal distress from before to after participating in the workshop. Furthermore, students’ trait empathy correlated negatively with the self-perceived sustainability of their solutions. These findings highlight the need for future work studying the role of empathy in encouraging a sustainable design mindset among designers.

The Theory of Inventive Problem Solving (TRIZ) method and toolkit provides a well-structured approach to support engineering design with pre-defined steps: interpret and define the problem, search for standard engineering parameters, search for inventive principles to adapt, and generate final solutions. The research presented in this paper explores the neuro-cognitive differences of each of these steps. We measured the neuro-cognitive activation in the prefrontal cortex (PFC) of 30 engineering students. Neuro-cognitive activation was recorded while students completed an engineering design task. The results show a varying activation pattern. When interpreting and defining the problem, higher activation is found in the left PFC, generally associated with goal directed planning and making analytical. Neuro-cognitive activation shifts to the right PFC during the search process, a region usually involved in exploring the problem space. During solution generation more activation occurs in the medial PFC, a region generally related to making associations. The findings offer new insights and evidence explaining the dynamic neuro-cognitive activations when using TRIZ in engineering design.

Urgent societal problems, including climate change, require innovation, and can benefit from interdisciplinary solutions. A small body of research has demonstrated the potential of positive emotions (e.g., gratitude, awe) to promote creativity and prosocial behavior, which may help address these problems. This study integrates, for the first time, psychology research on a positive and prosocial emotion (i.e., gratitude) with engineering-design creativity research. In a pre-registered study design, engineering students and working engineers (pilot N = 49; full study N = 329) completed gratitude, positive-emotion control, or neutral-control inductions. Design creativity was assessed through rater scores of responses to an Alternate Uses Task (AUT) and a Wind-Turbine-Blade Repurposing Task (WRT). No significant differences among AUT scores emerged across conditions in either sample. While only the pilot-study manipulation of gratitude was successful, WRT results warrant further studies on the effect of gratitude on engineering-design creativity. The reported work may also inform other strategies to incorporate prosocial emotion to help engineers arrive at more original and effective concepts to tackle environmental sustainability, and in the future, other problems facing society.

Previous efforts in the area of collaborative computer-aided design (CAD) suggest that a team of designers working synchronously in a multi-user CAD (MUCAD) environment can produce CAD models faster than a single user. Our research is the among the first to investigate assemblies in MUCAD. Due to the lack of hierarchical feature dependency in assemblies, we propose that CAD teams can optimize assembly through modularization and parallel execution. In our study, 20 participants were tasked with assembling pre-modelled CAD parts of varying complexity in teams of one, two, three or four. We analyze audio recordings, team activity, and survey responses to compare the performance of individuals and virtual collaborative teams during assembly, while working with the same MUCAD platform. This paper features a multimodal approach to analyze team trends in communication, workflow, task allocation and challenges to determine which factors are conducive to the success of a multi-user CAD team and which are detrimental. In our work, the success of a team is measured by its productivity score, which is the number of mates added by a team within a given time frame. We present evidence that teams can complete an assembly in less calendar time than a single user, but single users are more efficient based on person-hours, due to communications and coordination overheads. Surprisingly, paired contributors exhibit an assembly bonus effect. These findings represent a preliminary understanding of collaborative CAD assembly work. Our work supports the claim that collaborative assembly activities have the potential to improve the capabilities of modern product design teams, delivering products faster and at lower cost. We identify areas for future research, and highlight areas of improvement for collaborative CAD platforms and engineering design teams.

In an increasingly interconnected & cyber-physical world, the ability to coherently measure and manage complexity is vital for the engineering design and systems engineering community. To this end, numerous measures have been promulgated in the literature, yet these measures differ in terms of their intellectual foundations and perspectives, with limited cross-validation among them. In this paper, we propose a framework for benchmarking the status quo of existing complexity measurement approaches in terms of their alignment with the commonly-held beliefs in the literature. We discover that the literature broadly suggests an understanding of complexity based on a system’s size, number of interconnections, and architectural structure. We adopt a design of experiments approach and synthetically create system architectures to mimic the variation across these dimensions. We then use these architectures as a shared test-bed to document the response of four complexity measures that are representative of the predominant perspectives of the literature. We do this by evaluating the change in measurement of a complexity measure as we incrementally varied the levels of one system architecture property believed to affect complexity while keeping the others constant. We find that none of the measures fully satisfy the commonly-held beliefs of the literature and provide a discussion on the underlying factors that lead to these discrepancies. We note that multiple independent discussions coexist in the literature, with little cohesion and communication across the groups, suggesting that further research is required to understand the interactions and influences among these communities. For this purpose, our rigorous, structured, and grounded in literature benchmarking approach can serve as a testbed for development and verification of future architectural assessment tools and measures.

Knowledge organization is an essential component of engineering design, and a deeper understanding of how designers organize knowledge could enable more effective insights in support of the design process. To explore this, we examine 23 professional designers’ knowledge organization practices as they virtually engage with data collected during a teardown of a consumer product. Designers organized this data by forming groups of related data, nesting subgroups of data within groups, and creating directional links between groups of data and individual data. Our results indicate three insights about how designers organize and acquire knowledge from product teardowns. First, we observe that while designers find grouping data to be more effective for learning, linking proved more helpful for knowledge transfer. Second, we find that designers employ links between data much more frequently than they do nests, and that links primarily serve to identify trade-offs, requirements, and opportunities for team collaboration. Finally, a graph analysis indicates that design features, product housing, cost, and manufacturing coexist as separate but central groups in designers’ knowledge organization, reflecting the diversity of perspectives on knowledge organization emergent in a constrained teardown activity. These findings provide insight into professional designers’ knowledge organization practices, and represent a preliminary step toward design knowledge bases that more accurately reflect designer behavior, ultimately enabling more effective data-driven support tools for design.

As Artificial Intelligence (AI) assistance tools become more ubiquitous in engineering design, it becomes increasingly necessary to understand the influence of AI assistance on the design process and design effectiveness. Previous work has shown the advantages of incorporating AI design agents to assist human designers. However, the influence of AI assistance on the behavior of designers during the design process is still unknown. This study examines the differences in participants’ design process and effectiveness with and without AI assistance during a complex drone design task using the HyForm design research platform. Data collected from this study is analyzed to assess the design process and effectiveness using quantitative methods, such as Hidden Markov Models and network analysis. The results indicate that AI assistance is most beneficial when addressing moderately complex objectives but exhibits a reduced advantage in addressing highly complex objectives. During the design process, the individual designers working with AI assistance employ a relatively explorative search strategy, while the individual designers working without AI assistance devote more effort to parameter design.

AI technologies are enabling the development of not only active tools that provide decision-support, but also interactive tools that seek human input and feedback. As interactive tools facilitate human-AI interaction, their design needs to be informed by human-centric requirements, that is, the needs of the users of such tools. In the context of engineering design, there is a gap in our understanding of designing intelligent tools that facilitate human-AI interaction. To fill this gap, the research question of this study is, What are the human-centric design requirements for the design of AI agents to enable human-AI interaction in engineering design contexts? To answer this question, we conducted an interview study with faculty members in engineering design. The faculty predominantly discussed engineers, designers, and engineering design students as the potential stakeholders who would directly benefit from human-AI interaction. For such stakeholders, we identify several human-centric design requirements and challenges in designing AI tools that facilitate human-AI interaction in engineering design. We find that the requirements focused on the need to understand the stakeholders’ cognition and the engineering design contexts. The results of our study point to the need for the theory of mind in AI agents to enable them to infer stakeholder preferences while engaging in engineering design activities.

Nearly 60 years ago, Thomas Kuhn revolutionized how we think of scientific discovery and innovation when he identified that scientific change can occur in incremental developments that improve upon existing solutions, or it can occur as drastic change in the form of a paradigm shift. In engineering design, both types of scientific change are critical when exploring the solution space. However, most methods of examining design outputs look at whether an idea is creative or not and not the type of creativity that is deployed or if we can predict what types of individuals or teams is more likely to develop a paradigm-shifting idea. Without knowing how to identify who will generate ideas that fit a certain paradigm, we do not know how to build teams that can develop ideas that better explore the solution space. This study provides the first attempt at answering this question through an empirical study with 60 engineering design student teams over the course of a 4- and 8-week design project. Specifically, we sought to identify the role of cognitive style using KAI score, derived from Kirton’s Adaption-Innovation (A-I) theory, on the paradigm-relatedness of ideas generated by individuals and teams. We also sought to investigate the role of crowdsourcing for measuring the paradigm-relatedness of design solutions. The results showed that KAI was positively related to a greater likelihood of an individual’s idea being categorized as paradigm-breaking. In addition, the team KAI diversity was also linked to a greater likelihood of teams’ ideas being categorized as paradigm-challenging. Finally, the results support the use of crowdsourcing for measuring the paradigm-relatedness of design solutions.

Myoelectric signals from the human motor control system can improve the real-time control and neural-machine interface of robotic leg prostheses and exoskeletons for different locomotor activities (e.g., walking, sitting down, stair ascent, and non-rhythmic movements). Here we review the latest advances in myoelectric control designs and propose future directions for research and innovation. We review the different wearable sensor technologies, actuators, signal processing, and pattern recognition algorithms used for myoelectric locomotor control and intent recognition, with an emphasis on the hierarchical architectures of volitional control systems. Common mechanisms within the control architecture include 1) open-loop proportional control with fixed gains, 2) active-reactive control, 3) joint mechanical impedance control, 4) manual-tuning torque control, 5) adaptive control with varying gains, and 6) closed-loop servo actuator control. Based on our review, we recommend that future research consider using musculoskeletal modeling and machine learning algorithms to map myoelectric signals from surface electromyography (EMG) to actuator joint torques, thereby improving the automation and efficiency of next-generation EMG controllers and neural interfaces for robotic leg prostheses and exoskeletons. We also propose an example model-based adaptive impedance EMG controller including muscle and multibody system dynamics. Ongoing advances in the engineering design of myoelectric control systems have implications for both locomotor assistance and rehabilitation.

Heuristics are essential for addressing the complexities of engineering design processes. The goodness of heuristics is context-dependent. Appropriately tailored heuristics can enable designers to find good solutions efficiently, and inappropriate heuristics can result in cognitive biases and inferior design outcomes. While there have been several efforts at understanding which heuristics are used by designers, there is a lack of normative understanding about when different heuristics are suitable. Towards addressing this gap, this paper presents a reinforcement learning-based approach to evaluate the goodness of heuristics for three sub-problems commonly faced by designers: (1) learning the map between the design space and the performance space, (2) acquiring sequential information, and (3) stopping the information acquisition process. Using a multi-armed bandit formulation and simulation studies, we learn the suitable heuristics for these individual sub-problems under different resource constraints and problem complexities. Additionally, we learn the optimal heuristics for the combined problem (i.e., the one composing all three sub-problems), and we compare them to ones learned at the sub-problem level. The results of our simulation study indicate that the proposed reinforcement learning-based approach can be effective for determining the quality of heuristics for different problems, and how the effectiveness of the heuristics changes as a function of the designer’s preference (e.g., performance versus cost), the complexity of the problem, and the resources available.

This paper examines decision making under radical uncertainty in engineering design, that is, engineering decision making in those situations where it is not possible to know the outcomes and/or construct the utility functions and probabilities needed to support rational-human decision making. In these situations, despite being faced with radical uncertainty, engineers do (and must) proceed forward in a linear, clear, and predictable manner. Yet, they may not proceed in a manner that is well described by current engineering design frameworks. Examining the role of decision making in business and other social enterprises, Tuckett and Nikolic [1] have proposed conviction narrative theory (CNT) to describe how rational decision-makers confronted with situations in which insufficient information is available to support traditional decision-making tools use narrative and intuition to reach convincing and actionable decisions. This paper proposes that, in a manner similar to what is described in CNT, narrative and engineering judgment play a critical role in engineering design situations dominated by radical uncertainty. To that end, this paper integrates the traditional rational-human view of decision making as expressed by Hazelrigg in the well-known Decision-Based Design (DBD) framework and CNT as proposed by Tuckett and Nikolic. In the resulting rational, narrative-based design framework, narrative structures are used to describe and develop design alternatives and provide the ideas, beliefs, and preferences needed by the DBD framework. The resulting preferred design is expressed as a narrative and tested using engineering judgement. Specifically, the goal of the design process is expressed as a high-level guiding narrative that fosters the development of design narratives (design alternatives), and ultimately results in a convincing narrative that describes the preferred design. The high-level guiding narrative outlines the event(s), entity(s), preferences, and beliefs needed to support the design. The design narratives are narrative fragments that are nested within the high-level narrative and include the proposed action (idea), the specific challenges that the design faces, and the possible (but not yet verified) outcomes. The convincing narrative is the validated, preferred option that results from the DBD analysis and optimization process and is reviewed using engineering judgement. Following development of the rational, narrative-based design framework, the value of the framework is discussed within the context of practical engineering design.

Advances in additive manufacturing enable the fabrication of complex structures with intricate geometric details. It also escalates the potential for high-resolution structure design. However, the increasingly finer design brings computational challenges for structural optimization approaches such as topology optimization (TO) since the number of variables to optimize increases with the resolutions. To address this issue, two-scale TO paves an avenue for high-resolution structural design. The design domain is first discretized to a coarse scale, and the material property distribution is optimized, then using micro-structures to fill each property field. In this paper, instead of finding optimal properties of two scales separately, we reformulate the two-scale TO problem and optimize the design variables concurrently in both scales. By introducing parameterized periodic cellular structures, the minimal surface level-parameter is defined as the material design parameter and is implemented directly in the optimization problem. A numerical homogenization method is employed to calculate the elasticity tensor of the cellular materials. The stiffness matrices of the cellular structures derived as a function of the level parameters, using the homogenization results. An additional constraint on the level parameter is introduced in the structural optimization framework to enhance adjacent cellulars interfaces’ compatibility. Based on the parameterized micro-structure, the optimization problem is solved concurrently with an iterative solver. The reliability of the proposed approach has been validated with different engineering design cases. Numerical results show a noticeable increase in structure stiffness using the level parameter directly in the optimization problem than the state-of-art mapping technique.

The design of strongly coupled multidisciplinary engineering systems is challenging since it is characterized by the complex interaction of different disciplines. Such complexity cannot be easily captured by explicit analytical solutions, which motivates the development of surrogate modeling. It enables the prediction of the systems’ behavior without analytical formulations. Among existing surrogate modeling techniques, deep learning has gained significant interest because of the flexibility of non-linear formulation and applicability to data-driven analysis. Notably, the convolution neural networks-based deep surrogate model augments the precision of prediction and estimation of system behavior once image-based inputs representing physical experiments and simulation are employed. Nevertheless, the feasibility of the deep surrogate model is often flawed due to the miserable correlation representation between design parameters and the corresponding responses. Massive training costs also degrade the performance of the predictive model. To address those issues, this research proposes a physics-informed artificial image (PiAI) that incubates geometry-informed CAD, location-clarified filter, and essential simulation conditions, which augments the prediction credibility. Moreover, in lieu of employing multimodalities or multiple image channels, the proposed method employs a unimodal-based single image input to increase computational efficiency. The proposed framework’s efficacy and applicability are addressed in practical engineering design applications: cantilever beam and stretchable strain sensor.

In this study, the extractive summarization using sentence embeddings generated by the finetuned BERT (Bidirectional Encoder Representations from Transformers) models and the K-Means clustering method has been investigated. To show how the BERT model can capture the knowledge in specific domains like engineering design and what it can produce after being finetuned based on domain-specific datasets, several BERT models are trained, and the sentence embeddings extracted from the finetuned models are used to generate summaries of a set of papers. Different evaluation methods are then applied to measure the quality of summarization results. Both the automatic evaluation method like Recall-Oriented Understudy for Gisting Evaluation (ROUGE) and the statistical evaluation method are used for the comparison study. The results indicate that the BERT model finetuned with a larger dataset can generate summaries with more domain terminologies than the pretrained BERT model. Moreover, the summaries generated by BERT models have more contents overlapping with original documents than those obtained through other popular non-BERT-based models. It can be concluded that the contextualized representations generated by BERT-based models can capture information in text and have better performance in applications like text summarizations after being trained by domain-specific datasets.

Computational Fluid Dynamics (CFD) simulations are useful to the field of engineering design as they provide deep insights on product or system performance without the need to construct and test physical prototypes. However, they can be very computationally intensive to run. Machine learning methods have been shown to reconstruct high-resolution single-phase turbulent fluid flow simulations from low-resolution inputs. This offers a potential avenue towards alleviating computational cost in iterative engineering design applications. However, little work thus far has explored the application of machine learning image super-resolution methods to multiphase fluid flow (which is important for important for emerging fields such as marine hydrokinetic energy conversion). In this work, we apply a modified version of the Super-Resolution Generative Adversarial Network (SRGAN) model to a multiphase turbulent fluid flow problem, specifically to reconstruct fluid phase fraction at a higher resolution. Two models were created in this work, one with a simple physics-constrained loss function and one without, and the results are discussed and analyzed. We found that both models were able to significantly outperform non-machine learning upsampling methods and can preserve an impressive amount of detail and nuance, showing the versatility of the SRGAN model for upsampling fluid simulations. However, the difference in accuracy between the two models is quite minimal. This indicates that, for these contexts studied here, the additional complexity of a physics-informed approach may not be justified.