skip to main content


Title: Autonomous Manufacturing Using Machine Learning: A Computational Case Study With a Limited Manufacturing Budget
Abstract

This paper studies the concept of manufacturing systems that autonomously learn how to build parts to a user-specified performance. To perform such a function, these manufacturing systems need to be adaptable to continually change their process or design parameters based on new data, have inline performance sensing to generate data, and have a cognition element to learn the correct process or design parameters to achieve the specified performance. Here, we study the cognition element, investigating a panel of supervised and reinforcement learning machine learning algorithms on a computational emulation of a manufacturing process, focusing on machine learning algorithms that perform well under a limited manufacturing, thus data generation, budget. The case manufacturing study is for the manufacture of an acoustic metamaterial and performance is defined by a metric of conformity with a desired acoustic transmission spectra. We find that offline supervised learning algorithms, which dominate the machine learning community, require an infeasible number of manufacturing observations to suitably optimize the manufacturing process. Online algorithms, which continually modify the parameter search space to focus in on favorable parameter sets, show the potential to optimize a manufacturing process under a considerably smaller manufacturing budget.

 
more » « less
Award ID(s):
1727894
NSF-PAR ID:
10292306
Author(s) / Creator(s):
; ; ;
Date Published:
Journal Name:
International Manufacturing Science and Engineering Conference
Page Range / eLocation ID:
V002T07A009
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. The potential defects during the additive manufacturing (AM) process greatly deteriorate the mechanical properties of the fabricated structures and, as a result, increase the risks of part fatigue failure and even disasters. As laser additive manufacturing is such a complex process, many different physical phenomena such as electromagnetic radiation, optical and acoustic emission, and plasma generation will occur. Unlike vision and acoustic methods, the spectroscopy based smart optical monitoring system (SOMS) provides atomic level information revealing mechanical and chemical condition of the product. By monitoring plasma, multiple information such as line intensity, standard deviation, plasma temperature, or electron density, and by using different signal processing algorithms such as vector machine training or wavelet transforming, AM defects have been detected and classified. Utilizing two fiber optic components, a bifurcated fiber and a split fiber, the experimental results were performed to improve SOMS signal-to-noise ratio. Defects, including subsurface pores and sudden changes of process parameters including shielding gas shut-off and foreign substance, were identified by the spectroscopy based SOMS. For chemical composition characterization, a degree of dilution in terms of chemical element variation is identified by a spectral peak intensity ratio through the SOMS. It turned out that the information on the Cr/Fe ratio of deposit at a certain layer is vital to design the mechanical property in the IN625 deposition on the mild steel case. The SOMS has also demonstrated that the chemistry ratio can be determined from the calibration curve method based on the known alloy samples and that the ratio of the maximum intensities of multiple species provides more information about the quality of the alloy.

     
    more » « less
  2. Laser beam powder bed fusion (LB-PBF) is a widely-used metal additive manufacturing process due to its high potential for fabrication flexibility and quality. Its process and performance optimization are key to improving product quality and promote further adoption of LB-PBF. In this article, the state-of-the-art machine learning (ML) applications for process and performance optimization in LB-PBF are reviewed. In these applications, ML is used to model the process-structure–property relationships in a data-driven way and optimize process parameters for high-quality fabrication. We review these applications in terms of their modeled relationships by ML (e.g., process—structure, process—property, or structure—property) and categorize the ML algorithms into interpretable ML, conventional ML, and deep ML according to interpretability and accuracy. This way may be particularly useful for practitioners as a comprehensive reference for selecting the ML algorithms according to the particular needs. It is observed that of the three types of ML above, conventional ML has been applied in process and performance optimization the most due to its balanced performance in terms of model accuracy and interpretability. To explore the power of ML in discovering new knowledge and insights, interpretation with additional steps is often needed for complex models arising from conventional ML and deep ML, such as model-agnostic methods or sensitivity analysis. In the future, enhancing the interpretability of ML, standardizing a systemic procedure for ML, and developing a collaborative platform to share data and findings will be critical to promote the integration of ML in LB-PBF applications on a large scale. 
    more » « less
  3. null (Ed.)
    Utilization of multiple-input multiple-output (MIMO) systems as a means of increasing channel capacity has been an area of increasing consideration in radio communications. However, less study has been devoted to MIMO in the high-frequency band. This research is important because high-frequency communication using MIMO allows for international communication at long distances using lower power consumption than many other approaches. The inter-symbol interference caused by the selective fading of multiple received signals and the randomness of the ionospheric conditions means there is a need for a novel solution. The purpose of this research is to introduce two machine learning approaches that can adaptively apply equalization algorithms to address fading and optimize equalization parameters. The novelty of our approach lies in two main factors. The first is that our approach allows for a software-defined radio to switch equalization algorithms depending on conditions during run-time. The second is that we optimize this selected algorithm further by using two machine-learning approaches. The first proposed cognitive engine model, which utilizes a genetic algorithm, demonstrates the validity and advantage of using a cognitive engine to select optimal equalization parameters at the receiver under the problems created by utilizing the high-frequency band. This approach acts as a comparison for our second. We then propose a second cognitive engine, the adaptive manipulator, which optimizes not only by selecting equalization parameters but also continually changes the equalization algorithm. Finally, we compare the performance of the proposed cognitive engine models with state-of-the-art algorithms. 
    more » « less
  4. null (Ed.)
    Modern digital manufacturing processes, such as additive manufacturing, are cyber-physical in nature and utilize complex, process-specific simulations for both design and manufacturing. Although computational simulations can be used to optimize these complex processes, they can take hours or days--an unreasonable cost for engineering teams leveraging iterative design processes. Hence, more rapid computational methods are necessary in areas where computation time presents a limiting factor. When existing data from historical examples is plentiful and reliable, supervised machine learning can be used to create surrogate models that can be evaluated orders of magnitude more rapidly than comparable finite element approaches. However, for applications that necessitate computationally- intensive simulations, even generating the training data necessary to train a supervised machine learning model can pose a significant barrier. Unsupervised methods, such as physics- informed neural networks, offer a shortcut in cases where training data is scarce or prohibitive. These novel neural networks are trained without the use of potentially expensive labels. Instead, physical principles are encoded directly into the loss function. This method substantially reduces the time required to develop a training dataset, while still achieving the evaluation speed that is typical of supervised machine learning surrogate models. We propose a new method for stochastically training and testing a convolutional physics-informed neural network using the transient 3D heat equation- to model temperature throughout a solid object over time. We demonstrate this approach by applying it to a transient thermal analysis model of the powder bed fusion manufacturing process. 
    more » « less
  5. Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)
    The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should produce identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster. Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive. These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations. The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity. GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient. In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc. To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise. We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues. 
    more » « less