Table 1.

Definition of common terms used in AM and AI.

Common Terms Used in Additive Manufacturing and Artificial Intelligence	REFs
3D printing: three-dimensional (3D) printing is an additive manufacturing process in which a physical object is created from a computer-aided design (CAD) model by printing the model on a pre-computed layer-by-layer toolpath. This process is fully deterministic and, therefore, is ideal for printing on planar surfaces that are stationary relative to the coordinate system of the printer (namely, ex situ 3D printing). To date, there are several 3D printing methods that include the following: fused deposition modeling (FDM), selective laser sintering (SLS), stereolithography (SLA), and direct ink writing (DIW).	[1,18,24,28,40]
4D printing: four-dimensional (4D) printing uses the same techniques of 3D printing through computer-programmed deposition of material in successive layers to create a 3D object. However, in 4D printing, the resulting 3D object is able to change shape, structure, or function directly off the print bed in response to external stimulus, with the fourth dimension being the time-dependent shape change after the printing. It is therefore a type of programmable 3D printer, wherein after the fabrication process, the printed material reacts with parameters within the environment (humidity, temperature, mechanical force, pH, etc.) and changes its form accordingly.	[2,4,47,49,50,59,80]
Artificial Intelligence: artificial intelligence (AI) leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind. Although a number of definitions of AI have surfaced over the last few decades, the most used is that of John McCarthy: “it is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable”.	[56,58,87]
Machine Learning: machine learning (ML) is a branch of AI and computer science, which focuses on the use of data and algorithms to imitate the way those humans learn, gradually improving its accuracy. ML involves the development and deployment of algorithms that, rather than being programmed to assign certain outputs in response to specific inputs from the environment, analyze data and their properties, and determine the action by using statistical tools. Usually, ML algorithms can be broadly classified into the following five categories: supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning and federated learning.	[56,57,87]
Open-Loop AI printing: open-loop AI leverages pre-acquired sensory data (such as laser scanning and 3D tomography reconstructions) to obtain precise target geometry in various forms of 3D representations such as meshes and voxels. Then this geometry is calibrated with respect to the printing platform, thus enabling the generation of a toolpath on complex surfaces (i.e., organs or tissues). Based on this morphing path, open-loop AI can design the distribution of shape-morphing materials (whereby the morphing can be induced by mechanical load, change of temperature or pH, swelling) within the 3D-printed model to achieve improved compliance to a dynamically varying target surface. The AI-related computation occurs prior to the printing process.	[87,90]
Closed-Loop AI printing: closed-loop AI printing integrates sensing as part of the printing process. The sensory data are processed in real time using AI tools to recognize the surface of the target. A feedback-control system adjusts the toolpath in real time to compensate the target motion, environmental disturbance, and calibration errors, thus ensuring the 3D printing procedures.	[87,90,92]