AWS re:Invent 2020 looked quite different from past years––a three-week virtual event with remote access to keynotes and breakouts for leaders, engineers, partners, and others. Although the virtual format was different from the past, one thing was the same, the flagship event drew, engaged, and inspired thousands of attendees from around the world and launched new AWS offerings.
The conference’s remote nature opened the opportunity for more leaders, engineers, business developers, product owners, and marketing teams to participate. The IoT industry found ways to attend together remotely and discuss reactions, impressions, and takeaways, much like if we were attending in person.
IoT Is a Valuable Solution
Werner’s keynote featured a company called AVA that repurposed a wearable IoT device they already had in the field for tracking fertility into a COVID symptom tracker. Customers have proposed similar ideas about using products and systems already built as a foundation to help solve various issues presented by the pandemic. With current platforms assisting in proximity detection of people within six feet of one another, the enhancements for COVID were developed quickly and allows customers to hit the market during an impactful time of the pandemic. The opportunity for IoT to provide solutions for our world’s most challenging issues remains high for those who are willing to discover, innovate, and partner.
Machine Learning: Forging Ahead Expeditiously.
Machine Learning (ML) was a big part of re:Invent this year and certainly an area where AWS is investing heavily. As machine learning and artificial intelligence have advanced over the last few years, we have seen very rapid growth in machine learning support on cloud platforms. Because of this, most of the primary building blocks of machine learning, such as development support, deployment tools, framework flexibility, and pre-made models, have been developed and deployed. Now that those building blocks are in place, there seems to be a shift in new services moving to two recent big trends: MLOps and Turnkey ML Solutions.
Just as software development has seen a significant push for streamlining the full software development lifecycle through DevOps over the previous years, we are now seeing a similar push for streamlining the entire machine learning lifecycle through MLOps. AWS has been working to make it easy to plug their various ML services into a well understood and easy-to-use pipeline. Due to the complexity and cyclical nature of machine learning development, this will be a critical component of successful machine learning solutions. More details on the AWS implementation of this
Turnkey ML Solutions
Another area seeing a lot of progress is turn-key solutions for common machine learning problems. Gone are the days where you needed a dedicated team of data scientists, machine learning experts, and developers to get a solution up and running. A turnkey ML solution is a type of system built end-to-end for a customer that can be easily implemented into a current business process. It is immediately ready to use upon implementation and designed to fulfill a particular function. There is typically a trade-off with turnkey solutions, including cost, speed, or flexibility. While these solutions aren’t going to be a fit for everyone, companies can benefit significantly from them. As more common machine learning Applications get identified, we expect more of these services to launch.
IoT Continues To Advance Exponentially
AWS continues to lead with new and updated services that support the growth of IoT. These are just a few of the ones we feel are worth highlighting.
AWS provides a Long Term Support (LTS) release of the FreeRTOS operating system – a commonly used operating system in IoT devices. Amazon’s commitment to supporting this open-source operating system for security updates and critical bug fixes will simplify IoT products’ maintenance.
IoT Greengrass 2.0 is released, and it enhances the capabilities of managing fleets of devices. This new version is open-sourced and makes it simpler to debug problems with IoT systems, and lowers the cost of developing and managing the system. Read More
AWS IoT Core has now expanded to include long-range, low-power wide-area network (LoRaWAN) connectivity. This capability accelerates the development of connected systems where WiFi, Bluetooth, or cellular technologies are not a viable solution.
IIoT More Compelling for the Industrial Sector
A variety of new services targeting Industry 4.0 were launched and lower the entry barrier to IoT for the industrial sector. Although there is still custom work to be done, these new services make IIoT more welcoming. IoT SiteWise Edge is a service for collecting data from industrial equipment. Amazon Monitron is a product and service for monitoring equipment and enabling predictive maintenance. Amazon Lookout for Equipment is a targeted machine learning service for detecting abnormal equipment behavior.
No Equal Replacement for In-Person Learning and Networking
Amazon proved that it is possible to host huge conferences remotely during a pandemic. They did a fantastic job making the re:Invent conference remote, free, and accessible to thousands of attendees. It worked for this time; however, our team agreed that in-person learning creates a better focus on education, growth, and networking with other passionate software professionals. We hope to return to in-person attendance at learning events as soon as it’s safe to do so.
Amazon made the sessions available for free, so if you haven’t already checked them out, we encourage you to do so.