\n
\n
\n
\n
La inteligencia artificial en el borde se refiere al despliegue de algoritmos y modelos de IA directamente en dispositivos locales del borde, como sensores o dispositivos del Internet de las Cosas (IoT), lo que permite el procesamiento y an\u00e1lisis de datos en tiempo real sin depender constantemente de la infraestructura en la nube.<\/p>\n
What is edge AI?<\/strong><\/p>\n <\/a><\/div>\n Edge\u00a0artificial intelligence\u00a0refers to the deployment of\u00a0AI algorithms\u00a0and\u00a0AI models\u00a0directly on local\u00a0edge devices\u00a0such as sensors or\u00a0Internet of Things<\/a>\u00a0(IoT) devices, which enables\u00a0real-time data processing\u00a0and analysis without constant reliance on cloud infrastructure.<\/p>\n<\/div>\n<\/div>\n In essence, edge AI, or “AI on the edge\u201c, refers to the\u00a0combination of\u00a0edge computing<\/a>\u00a0and\u00a0artificial intelligence<\/a>\u00a0to perform\u00a0machine learning<\/a>\u00a0tasks directly on interconnected\u00a0edge devices.\u00a0Edge computing allows data to be stored close to the device location, and AI algorithms enable processing right on the network edge, with or without an internet connection. This capability facilitates data processing within milliseconds, providing real-time feedback.<\/p>\n Self-driving cars,\u00a0wearable\u00a0devices,\u00a0security cameras and smart home appliances are among the technologies that use edge AI\u00a0capabilities to promptly deliver users with\u00a0real-time\u00a0information when it is most essential.<\/p>\n Edge AI is becoming popular as industries find new ways to use its power to optimize workflows, automate business processes and foster innovation. At the same time, it helps address critical concerns like latency, security and cost reduction.<\/p>\n Learn more about IBM\u2019s\u00a0edge computing\u00a0solutions<\/a>.<\/p>\n Edge AI versus distributed AI<\/strong><\/p>\n<\/div>\n Edge AI enables onsite decision-making, eliminating the need to constantly transmit data to a central location and wait for processing, which streamlines the automation of business operations. However, data still needs to be transmitted to the cloud for retraining AI pipelines and deploying updated models.<\/p>\n Deploying this pattern across numerous locations and diverse applications presents challenges such as data gravity, heterogeneity, scale and resource constraints. Distributed AI helps overcome these obstacles by integrating intelligent data collection, automating the data and AI lifecycles, adapting and monitoring spokes and optimizing data and AI pipelines.<\/p>\n Distributed artificial intelligence (DAI) is responsible for distributing, coordinating and forecasting task, objective or decision performance within a multiagent environment.\u00a0DAI scales applications across numerous spokes and enables AI algorithms to autonomously process across multiple systems, domains and devices on the edge.<\/p>\n<\/div>\n<\/div>\n Edge AI versus cloud AI<\/strong><\/p>\n<\/div>\n Presently,\u00a0cloud computing<\/a>\u00a0and application programming interfaces (APIs) are used to train and deploy\u00a0machine learning models. Later,\u00a0edge AI\u00a0conducts\u00a0machine learning\u00a0tasks such as predictive analytics, speech recognition and anomaly detection close to the user, distinguishing itself from the common\u00a0cloud services\u00a0in various ways. Instead of applications being developed and run entirely on the cloud,\u00a0edge AI systems\u00a0process and analyze data closer to the point where it was created.<\/p>\n Machine learning algorithms\u00a0are able to run on the edge and information can be processed right onboard\u00a0IoT devices, rather than in a private\u00a0data center<\/a>\u00a0or in a\u00a0cloud computing\u00a0facility.<\/p>\n Edge AI\u00a0presents itself as a better option whenever\u00a0real-time\u00a0prediction and\u00a0data processing\u00a0are required. Consider the most recent\u00a0advancements\u00a0in self-driving vehicle technology.\u00a0To ensure the secure navigation of these cars and their avoidance of potential dangers, they must rapidly detect and respond to a range of factors such as traffic signals, erratic drivers and lane changes. In addition, they must account for pedestrians, curbs and numerous other variables.<\/p>\n Edge AI\u2019s ability to locally process this information within the vehicle mitigates the potential risk of\u00a0connectivity problems that might arise from sending data to a remote server through\u00a0cloud-based\u00a0AI. In scenarios of this nature, where quick data responses could determine life or death outcomes, the vehicle’s ability to react swiftly is crucial.<\/p>\n Conversely, cloud AI refers to the deployment of\u00a0AI algorithms\u00a0and models on cloud servers. This method offers increased\u00a0data storage\u00a0and processing power capabilities, facilitating the training and deployment of more advanced\u00a0AI models.<\/p>\n<\/div>\n<\/div>\n Key differences between edge AI and cloud AI<\/strong><\/p>\n<\/div>\n Cloud AI can provide greater computational capabilities and storage capacity compared to edge AI, facilitating the training and deployment of more intricate and advanced AI models. Edge AI’s processing capacity is limited by the device’s size constraints.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n Latency<\/a>\u00a0directly affects productivity, collaboration, application performance and user experience. The higher the\u00a0latency\u00a0(and the slower\u00a0response times) the more these areas suffer.\u00a0Edge AI provides reduced\u00a0latency\u00a0by processing data directly on the device, whereas cloud AI involves sending data to distant servers, leading to increased\u00a0latency.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n Bandwidth\u00a0refers to the public data transfer of inbound and outbound network traffic around the globe.\u00a0Edge AI\u00a0calls for lower\u00a0bandwidth\u00a0due to local\u00a0data processing\u00a0on the device, whereas cloud AI involves data transmission to distant servers, demanding higher network\u00a0bandwidth.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n Edge\u00a0architecture\u00a0offers enhanced privacy by processing\u00a0sensitive data\u00a0directly on the device, whereas cloud AI entails transmitting data to external servers, potentially exposing sensitive information to third-party servers.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n Benefits of edge AI for end users<\/strong><\/p>\n <\/a><\/div>\n In 2022, the global\u00a0edge AI\u00a0market was valued at USD 14,787.5 million and is expected to grow to USD 66.47 million by the year 2023, according to a report conducted by\u00a0Grand View Research, Inc<\/span><\/a>. The rising demand for IoT-based edge computing services, along with edge AI’s inherent advantages, drives the rapid expansion of edge computing. The primary\u00a0benefits of edge AI\u00a0include:<\/p>\n<\/div>\n<\/div>\n Diminished latency<\/strong><\/p>\n<\/div>\n Through complete on-device processing, users can experience rapid response intervals without any delays caused by the need for information to travel back from a distant server.<\/p>\n<\/div>\n<\/div>\n Decreased bandwidth<\/strong><\/p>\n<\/div>\n As\u00a0edge AI processes\u00a0data on a local level, it minimizes the\u00a0amount of data\u00a0transmitted over the internet, leading to the preservation of internet\u00a0bandwidth. When less bandwidth\u00a0is used, the data\u00a0connection\u00a0can handle a larger volume of simultaneous data transmission and reception.<\/p>\n<\/div>\n<\/div>\n Real-time analytics<\/strong><\/p>\n<\/div>\n Users can perform\u00a0real-time data processing\u00a0on devices without the need for system\u00a0connectivity\u00a0and integration, enabling them to save time by consolidating data without needing to communicate with other physical locations.\u00a0However, edge AI might struggle to handle the vast volume and diversity of data required by certain AI applications.\u00a0To overcome these limitations, it must integrate with cloud computing\u00a0to use its resources and capabilities.<\/p>\n<\/div>\n<\/div>\n Data privacy<\/strong><\/p>\n<\/div>\n Privacy is increased because data is not transferred over to another network, where it becomes vulnerable to cyberattacks.\u00a0Through processing information locally on the device,\u00a0edge AI\u00a0reduces the risk for the mishandling of data. In industries subject to data sovereignty regulations,\u00a0edge AI\u00a0can aid in maintaining compliance by locally processing and storing data within designated jurisdictions.<\/p>\n However, any centralized database has the potential to become an enticing target for potential attackers,\u00a0meaning edge AI is still exposed to security risks.<\/p>\n<\/div>\n<\/div>\n Scalability<\/strong><\/p>\n<\/div>\n Edge AI\u00a0expands systems by using\u00a0cloud-based\u00a0platforms and inherent edge capabilities on original equipment manufacturer (OEM) technologies, encompassing both software and hardware. These OEM companies have begun to integrate native edge capabilities into their equipment, making it easier to scale the system. This expansion also enables local networks to maintain functionality even in situations where nodes upstream or downstream experience downtime.<\/p>\n<\/div>\n<\/div>\n Reduced costs<\/strong><\/p>\n<\/div>\n Expenses associated with AI services hosted on the cloud can be high.\u00a0Edge AI\u00a0offers the option of using costly cloud resources as a repository for postprocessing data accumulation, intended for subsequent analysis rather than immediate field operations. This reduces the workloads of\u00a0cloud computers\u00a0and networks.<\/p>\n The use of\u00a0CPU,\u00a0GPU\u00a0and memory experiences a large reduction as their workloads are distributed among\u00a0edge devices, distinguishing\u00a0edge AI\u00a0as the more cost-effective option between the two.<\/p>\n When\u00a0cloud computing\u00a0handles all the computations for a service, the centralized location bears a significant workload. Networks endure high traffic to transmit data to the central source. As machines execute tasks, the networks become active once more, transmitting data back to the user.\u00a0Edge devices\u00a0remove this continuous back-and-forth data transfer. As a result, both networks and machines experience reduced stress when they are relieved from the burden of handling every aspect.<\/p>\n Moreover, the\u00a0autonomous\u00a0traits of\u00a0edge AI\u00a0eliminate the need for continuous supervision by\u00a0data scientists.\u00a0Although human interpretation play a pivotal role in determining the ultimate value of data and the outcomes that it yields, edge AI platforms assume some of this responsibility. This shift ultimately leads to cost savings for businesses.<\/p>\n<\/div>\n<\/div>\n How does edge AI technology operate?<\/strong><\/p>\n <\/a><\/div>\n Edge AI\u00a0uses\u00a0neural networks<\/a>\u00a0and\u00a0deep learning<\/a>\u00a0to train models to accurately recognize, classify and describe objects within the provided data. This training process usually uses a centralized\u00a0data center\u00a0or the cloud to process the substantial volume of data necessary for model training.<\/p>\n After deployment,\u00a0edge AI models\u00a0progressively improve over time. When the AI encounters an issue, the problematic data is often transferred to the cloud for further training of the initial AI model, which ultimately replaces the inference engine at the edge.\u00a0This feedback loop significantly contributes to enhancing model performance.<\/p>\n<\/div>\n<\/div>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nComputing power<\/strong><\/p>\n\n\n\nLatency<\/strong><\/p>\n\n\n\nNetwork bandwidth<\/strong><\/p>\n\n\n\nSecurity<\/strong><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n