From Web to Machine: Amazon Industrial (AI) Services
Posted by Doug Harp
Business Insider reported that Amazon has filed a patent application for new software that will be used for monitoring industrial plants. The project is named Thor, and the software is intended to collect data from production, provide monitoring, analysis, predictions, and forecasts. Thor looks to provide Predictive Maintenance capabilities, using analysis of temperature, sound and vibration measurements to predict when a machine needs maintenance before a breakdown occurs. The Amazon AI services are currently being tested in Sweden and is expected to be formally launched in October. Initial uses target smart building management for large industrial facilities.
Amazon Web Services generated nearly $11 billion in revenues in 2Q20, representing 29% growth over the prior year. AWS has been a key factor in Amazon’s growth, and has transformed the landscape of enterprise computing. Launched in 2006, AWS has expanded its market footprint by aggressively leveraging its unique scale in order to deliver low-cost services to customers – and in 2Q20 AWS lowered prices again for its EC2 compute instances, continuing a trend of delivering increasing performance at lower and low prices.
The move into Industrial IoT services reflects a focus on higher value software offerings, where AWS can charge a premium while core compute and storage offerings help expand customers and usage with aggressive pricing. Amazon’s strategy for AWS has been to go after high margin enterprise technology vendors from the beginning (“Your margin is my opportunity”) – as the company commanded major spending share gains from traditional enterprise server and storage vendors including HP, IBM, Sun/Oracle, Dell and others. To date, only Microsoft has been able to challenge AWS as a cloud computing services platform at the enterprise, with Google a distant third.
Amazon Web Services has built a robust portfolio of offerings that span infrastructure, tools and applications, and its IoT offerings provide building blocks for a broad range of use cases as well as Amazon AI services. Much of AWS’s innovation around IoT has focused on connectivity and developer tools. AWS IoT Core enables users to connect devices with cloud applications, keep track of messages while processing and routing data to the appropriate endpoints. Greengrass is an innovative platform that enables local data processing by edge devices. Other offerings include Amazon AI services, such as FreeRTOS, SiteWise,Device Management and Things Graph provide a robust suite of capabilities for developers to create applications for monitoring industrial data from edge devices, sensors and other sources, employing advanced Amazon AI services and analytics to deliver predictive capabilities.
The competitive landscape is getting more crowded, as Microsoft, Google and scores of other providers offer software and cloud services that target the industrial IoT market. Tools do over time become commodified, and it’s the applications that provide the potential to generate higher margin, sticky revenue streams. There are many vendors and system integrators already offering industrial IoT solutions built on AWS, as well as a bevy of startups and traditional industrial firms targeting the market with specialized and bespoke offerings.
For Amazon, Thor represents a move “up the stack” into applications, where AWS’s successes have been mixed, with offerings like email, Chime (video conferencing) and calendaring failing to gain much traction against entrenched incumbents. In an emerging area like Predictive Maintenance, and other related domains around Industrial IoT, much of the competition consists of specialized, vertical or bespoke applications and the market is fragmented. With the broad footprint of AWS uses, Thor could have a good opportunity to establish meaningful share in the industrial market, and potentially expand use cases and usage overall.