MCD

Explore

We introduce Multi-Objective Counterfactuals for Design (MCD), a novel method for counterfactual optimization in design problems. MCD provides helps users reimagine their design concepts by recommending targeted modifications. Users can ask questions like: What if my design were lighter? What if my design were more futuristic? MCD reimagines design concepts in these counterfactual scenarios.

Code

Paper

15 Minute Presentation

Interactive Demo

Pypi Package

Overview

Users query MCD with an existing design concept and a set of desired counterfactual attributes. MCD iteratively queries a set of forward models (ML predictors, simulations, equations, etc.) provided by the user to identify a modification to the query that results in the desired counterfactual attributes.

MCD then recommends a diverse set of design modifications. These design modifications ideally display the counterfactual attributes while being as similar to the original design as possible.

Bike Frame Counterfactuals with Improved Mass and Safety Factor

In this case study, MCD was asked to modify an existing bike frame to make it nearly 30% lighter and 20% stronger without changing its material.

Using a predictive model trained on the FRAMED dataset, MCD was able to identify targeted modifications the the query design, which is represented across 37 continuous and categorical parameters.

MCD identifies numerous designs that achieve the performance targets. It makes modifications to a number of parameters, but consistently increases the thickness of the down tube, a key structural shortcoming of the original design. By adding material in the down tube, MCD is able to remove weight from other components, achieving the desired weight savings.

Querying MCD with parametric Counterfactual Attributes

In many design settings, designers may have subjective requirements that are difficult to quantify. We demonstrate that MCD can be queried using text-based and image-based counterfactual attributes. To allow us to calculate the match between a design and a text/image attribute, we first render an image of each design candidate, then calculate a text-to-image or image-image similarity score using a Contrastive Language Image Pretraining (CLIP) model.

In this image, we show a map of designs recommended by MCD when given a red road bike as an input design and a text prompt and reference image as counterfactual attributes. Similarity to the input design is prioritized at the top of the map and relaxed towards the bottom, making designs near the top closely resemble the original bike, but allowing designs towards the bottom to differ more significantly. To the left, similarity to the prompt: 'A futuristic black cyberpunk-style road racing bicycle' is prioritized. Towards the right, similarity to an image of a blue Fuji Wendigo 1.1 mountain bike is prioritized.

Querying MCD with Multimodal Counterfactual Attributes

Next, we demonstrate a more complex case where MCD is given six multimodal counterfactual attributes: the safety factor, weight limit, a text prompt, and image from the previous two examples, and two additional attributes: aerodynamic drag and ergonomic fit. Since aerodynamics and ergonomics are highly dependent on rider positioning and bike fit, we select a specific set of biometric parameters to design around, representing a sample rider. This example showcases MCD's ability to make individualized design customization.

The pairplot above shows the distribution of the six counterfactual attributes across the recommended designs. The diagonal shows the distribution of each attribute, while the off-diagonal shows the relationship between each pair of attributes. The designs recommended by MCD are diverse, with some designs prioritizing weight and safety, others aerodynamics and ergonomics, and others a balance of all attributes. However, all designs are within the bounds of the counterfactual attributes. We also denote the existing designs in the dataset, of which none meet all the counterfactual attributes.

MCD's Computational Cost

MCD is lightweight, contributing just 5% to the total optimization time in the previous multimodal counterfactual search.

We anticipate that MCD should be a lightweight modification to most design optimization pipelines.

Implicit Constraint Satisfaction

Compared to optimization, which struggles to satisfy design constraints that aren't explicitly specified, MCD can much more reliably satisfy design constraints implicitly.

While MCD generates bikes that are mostly feasibly, pure NSGA-II optimization generates designs with a variety of obvious physical infeasibilities, such as overlapping components.

Citations

ACKNOWLEDGEMENT

We would like to thank Amin Heyrani Nobari for his contributions to the image rendering pipeline that enabled much of the cross-modal work presented. We would also like to thank Tyler Butler for his feedback and edits.