Artificial intelligence (AI) is going mainstream, which has long-lasting implications for how enterprises build and improve their products and services.
As announced at re:Invent 2016, AWS is simplifying artificial intelligence (AI) adoption through three low-cost, cloud-based services built for AI-specific uses cases. Instead of creating proprietary algorithms, data models or machine learning techniques, all levels of developers from Global 2000 enterprises through start-ups can leverage the Amazon Lex, Amazon Rekognition and Amazon Polly APIs to innovate quickly and build new Internet of Things (IoT) product and service offerings. Accenture is delivering these innovative offerings by supporting clients with vertical industry applications powered by Amazon AI.
Combining AI with IoT is essential because it enables businesses to collect data in the physical world—from wearables, appliances, automobiles, mobile phones, sensors and other devices—and add intelligence to deliver a better response or outcome. In other words, AI is the automation brainpower to make IoT device-driven data more useful.
For example, a telecommunications company could create an AI-powered mobile chat bot to automate customer service processes. One use case would be to monitor incoming IoT data from cable boxes installed in homes. If a device started to malfunction, the mobile chat bot could notify the customer via text or voice interaction of a possible service issue, and offer the convenience of scheduling a service technician. This device-driven data could leverage AWS Lambda for serverless functions, as well as AWS Greengrass for embedded software on the edge. Thereby, leveraging the use of AWS cloud as needed when processing, storing and computing.
API functionality overview and real-world uses
Used separately or in combination, developers can embed the APIs into existing smart product and service roadmaps, or inject them into cloud-native programming processes.
Amazon Lex—AI-driven processing engine that computes voice input or sensor data to better understand and personalize an experience or outcome (part of Alexa voice platform)
Amazon Rekognition—Image and facial analysis to detect and understand environment and what is happening in real-life scenario or picture
Amazon Polly—Text-to-speech service that synthesizes structured text data into natural voice-like capability (in male or female voice and in 24 languages) to enrich response.
Today, businesses typically run analytics in the cloud on transactional datasets, such as customer purchases or location-based information. But IoT data combined with AI provides a deeper level of insights. By collecting real-time data from IoT devices (or what is known as device-driven data), a business can use an AI engine to automate the information processing and connect different sources of unstructured/structured data to contextualize what a person is asking for. From this understanding, the machine can provide a personalized response or experience directly to the end user, or route the response back into the enterprise to automate another process.
This capability opens an entirely new set of IoT-based product or service offerings. Accenture recently released their Technology Vision 2017, which explains the benefits of AI for the Enterprise. For instance, a healthcare business could implement Amazon Lex and Amazon Rekognition to improve the process of monitoring house-bound or elderly patients who need assisted living. In one use case, the service could install a video camera to take pictures of an individual, analyze the images in the cloud to keep track of movements, and send an automatic alert to a healthcare giver or family member if the patient has not moved in a specified amount of time or has fallen.
Expanding AI and IoT opportunities
In the future, AI combined with IoT will introduce even more scenarios in which robots (aka automated machines) collaborate with people to supply intelligent information and augment human interaction. This will help people to complete tasks more efficiently, interact in a more personalized way or supply on-demand services.
In a retail setting, for example, a business could create a collaborative artificial intelligence (“cobot”) application using Amazon Lex and Amazon Rekognition that analyzes facial features of in-store shoppers in real-time and combines this information with purchase transaction history. The cobot could then prompt sales associates to offer customized help to each customer as they choose items. Or in a hospital situation, Amazon Lex and Amazon Rekognition could be built into an application that uses AI and cloud-based big data, all connected with IoT, to help physicians better diagnose their patients. Examples include detecting skin anomalies with image analysis or stress-related symptoms.
With AWS’s new AI-driven APIs, developing IoT products and services with AI capabilities is becoming cost effective and accessible for all businesses--and leveraging Accenture to deliver new, applied solutions give enterprises a quick way to adopt scale.