A smart component can notify a manufacturer that it has reached the end of its life or is due for inspection. Rather than monitoring these data points externally, the part itself will check in occasionally with AI systems to report normal status until conditions go sideways, when the part will start demanding attention. This approach cuts down on the volume of data traffic within the system, which at scale can become a significant drag on analytic processing performance. Generative design can create an optimal design and specifications in software, then distribute that design to multiple facilities with compatible tooling. This means smaller, geographically dispersed facilities can manufacture a larger range of parts. These facilities could be proximal to where they’re needed; a facility might make parts for aerospace one day and the next day make parts for other essential products, saving on distribution and shipping costs.
Generative AI, a subset of AI, involves algorithms that can generate new content or designs from scratch, given a set of rules and inputs. It’s much like a skilled artist given a canvas, colors, and a general theme, who then creates an entirely new piece of art. In the context of manufacturing, this implies the creation of optimized design alternatives for parts, products, or even entire production processes.
2 Artificial Intelligence-Based Prognosis for Predictive Maintenance.
Software powered by artificial intelligence can help businesses optimise procedures to maintain high production rates indefinitely. To locate and eliminate inefficiencies, manufacturers may use AI-powered process mining technologies. Commonly known as “industrial robots,”robotics in manufacturingallow for the automation of monotonous operations, the elimination or reduction of human error, and the reallocation of human labour to higher-value activities. The convergence of AI, particularly generative AI, with metaverse and web3 technologies is creating a new frontier in manufacturing and industrial operations. Companies embracing this trinity of technologies will likely find themselves at the forefront of the next industrial revolution, armed with tools that foster innovation, efficiency, and sustainability. Predictive analytics can help anticipate demand patterns and optimize inventory management, while natural language processing can assist in automating customer service.
Recently, secondary battery has gained considerable interest worldwide due to its rising demand for electric vehicles (EVs) and hybrid electric vehicles (HEVs). One of the most commonly adopted secondary batteries for such vehicles is the lithium-ion battery because of its high power density, long battery life, high durability, low self-discharge rate, and fast charge rate compared to other types of secondary cells. The former is performed ex-situ by predicting in advance the battery behavior under specific conditions, while the latter is conducted in-situ by the battery management system (BMS) contained in a battery pack for EVs and HEVs. Autonomous driving (AD) is a thriving field of study where AI is actively taking part. The main objectives of AD consist of road detection, lane detection, vehicle detection, pedestrian detection, drowsiness detection, collision avoidance, and traffic sign detection [9].
Intel beats expectations as margins rise, manufacturing momentum builds
Even though image-based detection is taking over much of the highlights, there is a substantial amount of studies regarding signal-based methods. Lee et al. [86] showed that fault diagnosis to find root causes of process failures could be effectively carried out even using a black box CNN model. This is enabled particularly by tailoring the CNN’s receptive field over multivariate sensor signals along the time axis that allows for the association of its extracted features from hidden layers with the physical meaning of raw data. Lee et al. [87] focused on reducing the noise while maintaining valuable information as much as possible for reliable and robust fault monitoring. For reducing the noise, the author proposes SDAE for which several DAEs are pre-trained with latent representation from the previous time step given as input.
In human–robotic systems with limited interaction, a human supervisor will need to alter the system operation by overriding with manual controls or by stopping and re-teaching the system to prevent reoccurrence of the fault. An important aspect of intervention in the control of the HRC system is the capability for real-time decision-making. In complex and ambiguous situations, human operators have a relative advantage compared to robots owing to human cognitive abilities by virtue of experience.
Quality Control and Defect Detection
The images captured in unstable lighting conditions are pre-processed with Laws and Sobel filters to extract features, which are then fed to an SVM classifier enhanced by pyramid analysis. The proposed technique reaches a texture classification accuracy of 98% while satisfying the computation time requirements in a massive production setting. Prognosis aims at predicting the temporal progression of machine performance degradation, from its current state to final functional failure. In general, AI-based prognosis is part of the data-driven method that relies on establishing a machine performance evolution model to predict future machine performance based on its current and past status. To estimate RUL, one-step-ahead prediction is iteratively carried out until the predicted value passes a failure threshold [97]. In case that the machine performance is difficult to measure directly, an artificial health index (HI) is often created from sensor data to represent the machine performance [115].
Pan et al. [78] suggested an advanced neural network-based coating weight control approach for hot-dip galvanizing lines. The framework consisted of a feedforward control (FFC) and feedback control (FBC), together with a neural network predictive model, a bias-update module, and a real-time optimizer. Through this framework, nonlinearity, large time-variant delays, disturbances, and unsynchronized regulation of two manipulated variables (MVs) have been addressed. Both the coating weight variance and the transition time were greatly reduced as well. Mao et al. [79] introduced a groundbreaking neural network model consisting of the BP algorithm and the genetic algorithm for the first time to model and predict the thickness of the hot-dip galvanized zinc sheet.
How AI can democratize production of and access to goods
To imitate such patterns, the branch of visual CNN is fed with real-world images of an object, while the haptic CNN branch is fed with signals of five types of physical quantities (e.g., fluid pressure and core temperature.). The proposed model shows a high classification accuracy of objects initially labeled as 24 different haptic adjectives (e.g., bumpy, soft, porous, compressible, sticky, and textured). Polydoros et al. [46] proved the superiority of deep learning models in the learning of inverse dynamics of a robotic manipulator. The study suggests replacing the conventional physics-based models that cannot cope with the change in robot structure and dynamic environments with the proposed model, which is largely a DNN with one hidden layer modeled as an RNN. Similarly, Lenz et al. [47] used a deep learning-based framework called DeepMPC to handle robotic food-cutting, wherein a deep recurrent model is devised to model a time-varying nonlinear dynamics involved in the task.
- Machine learning solutions can promote inventory planning activities as they are good at dealing with demand forecasting and supply planning.
- Another key need in advancing HRC is being able to understand and learn the wide range of activities performed by the human operator.
- This is crucial for manufacturers to adjust production levels, resource allocation, and inventory management.
- AI-enabled energy management systems monitor energy consumption in real-time, identifying opportunities for optimizing and reducing energy waste.
- The alternating convolution and sub-sampling operations are first conducted at CNN, and then a generalized multi-layer network is eventually implemented.
- Watch this video to see how gen AI improves customer service for an automotive manufacturer, delivering real-time support to the vehicle owner who sees an unexpected warning light.
- That there remain significant data, and problem formulation challenges to be solved does not limit the already demonstrated opportunity for AI to transform manufacturing as we know it today.
But only 30 of them have been able to scale AI and other emerging technologies to drive business value. The extreme price volatility of raw materials has always been a challenge for manufacturers. Businesses have to adapt to the unstable price of raw materials to remain competitive in the market. AI-powered software like what is AI in manufacturing can predict materials prices more accurately than humans and it learns from its mistakes. The COVID-19 pandemic also increased the interest of manufacturers in AI applications. As seen on Google Trends graph below, the panic due to lockdowns may have forced manufacturers to shift their focus to artificial intelligence.
How Web3 Will Transform Industry
Companies are beginning to employ generative AI in their design and development stages. By feeding parameters and requirements into generative design software, companies can obtain optimized design solutions that not only meet their criteria but also present options they might not have considered. These designs can then be tested and refined in the metaverse, leading to innovative and efficient real-world applications.
It has been used to create new types of components that are cheaper, lighter, and sturdier than existing components, improving the overall qualities of many products from cars and aircraft to prefabricated houses and structures. Datamation is the leading industry resource for B2B data professionals and technology buyers. Datamation’s focus is on providing insight into the latest trends and innovation in AI, data security, big data, and more, along with in-depth product recommendations and comparisons.
Top 7 Data Modeling Tools You Need to Know in 2023
An interdisciplinary program that combines engineering, management, and design, leading to a master’s degree in engineering and management. Bring a business perspective to your technical and quantitative expertise with a bachelor’s degree in management, business analytics, or finance. Here are 11 innovative companies using AI to improve manufacturing in the era of Industry 4.0. It’s painful and expensive to migrate once you have all your data in a single cloud provider. Intel reported adjusted profits of 41 cents per share in the third quarter, compared to an estimate of 22 cents according to LSEG data. The company forecast adjusted current-quarter revenue of about $14.6 billion to $15.6 billion, compared with an estimate of $14.35 billion according to LSEG data.
Gen AI can play a key role in transforming maintenance workflows and staying one step ahead with predictive maintenance. It helps manufacturers optimize operations by interpreting telemetry from equipment and machines to reduce unplanned downtime, gain operating efficiencies, and maximize utilization. If a problem is identified, gen AI can also recommend potential solutions and a service plan to help maintenance teams rectify the issue. Manufacturing engineers can interact with this technology using natural language and common inquiries, making it accessible to the current workforce and attractive to new employees. Despite the varying degrees of applicability and research gaps that exist and need to be overcome in each of the four domains, the trend is undeniably one of the increased implementation of AI-based analytical tools.
The wavelet time-frequency spectra of vibration signals are first analyzed for energy-related feature extraction. Then, each signal was encoded by visual words or feature clusters, which were used as input to a sparse classifier to determine bearing fault type. The sparse classifier has also been investigated for gearbox fault severity level recognition [102]. The contribution of this work also includes a novel multi-sensor fusion method based on the covariance matrix, which allows pair-wise correlation among sensing signals to be estimated and incorporated into the analysis. In both works, the authors reported that the diagnosis accuracy from a sparse classifier is comparable to an SVM with reduced computational time. Sparse classifiers have also been investigated for wind turbine condition monitoring and fault diagnosis [103].