Interview: Dave Friedman Talks About How IoT Cloud Platforms Help to Accelerate Connected Product Delivery
Everyone is thinking about how the Internet of Things (IoT) can be applied to their products to automate operations and deliver new value to customers. But at the same time, many developers tasked with building connected products are struggling to assimilate all the new technologies involved and with how to craft an end-to-end architecture that is easy for customers to use, secure, scale, and process all the valuable data that the IoT can create. To help with all of this, engineering teams are now looking to application enablement platforms that are end-to-end and cloud based, such as the IoT Platform offered by Ayla Networks.
I talked with Dave Friedman, Ayla Networks’ Chief Executive Officer and Co-Founder, about the market evolution and acceptance of IoT Cloud Platforms from providers like Ayla.
This file type includes high resolution graphics and schematics when applicapable.
Wong: What challenges do engineering teams have in migrating to their current products to the Internet-connected products?
Friedman: Today, one of the big challenges is that engineering teams considering this sort of activity are typically looking at a complete re-architecture of their existing products. Most “things” that will connect to the Internet have simple, lightweight architectures, with no extra code space for all the additional code required for networking. Of course, the system re-architecture is just the start. Engineering teams are expected to somehow master all of the intricacies of networking, security stacks and certificates, fire walls, and data management. Add to that building out a scalable back-end that can manage the massive amount of data the devices will create, and management tools to provision and control the devices, and what you have is a recipe for failure for teams that try to do this themselves without a very big investment in time and money to hire and maintain a team that knows how to do this.
Related Articles
- Understanding Wireless Routing For IoT Networks
- How Many Quarks Does It Take To Make An IoT?
- The Internet Of Thingamajigs
Wong: I assume the challenges don’t end with the completion of the initial product development. Products must be supported and enhanced as well.
Friedman: That is really the key challenge. It is one thing to connect a device. But this is a new market, and so marketing will want to change the features of the product as they learn more about customer requirements. Engineering teams need to create a path that enables this sort of flexibility: products need to be able to change features quickly, and they need the entire platform to be able to manage all sorts of data types. In other words, engineering absolutely must build-in “future proofing” to enable marketing to modify the products quickly.
Beyond that, there is the issue of supporting the products that are in the field. There are firmware updates that need to be managed, ongoing maintenance of the backend to maintain high availability. Connected products will also create new challenges for customer support, so the engineering team needs to consider how to very quickly troubleshoot and resolve any connectivity problems. Eventually, customers will vote with their wallets, and purchase products that deliver a great experience.
Wong: What is an IoT Cloud Platform?
Friedman: Relevant for Ayla’s approach, we use the word “Platform” because we essentially provide a distributed set of software that works really well together. We provide drivers for microcontrollers, so our customers don’t have to re-architect their systems to work with our platform. We provide all of the device-side networking to enable the products to connect to our cloud service, so our customers don’t have to write any networking software of their own – and end users get a great experience. We provide the entire backend in the cloud as well. That means for every “tangible device” that is deployed with our service, we have a virtual instance defined in the cloud, and we manage that device, the data it creates, and the various credentials required to operate and control the device. Finally, we provide smart application libraries that dramatically ease development of great mobile applications.
So this is a “Platform” because there are so multiple components that work in different ways to provide a great experience for manufacturers and end users. The “nerve center” of the platform is in the cloud, and so that’s why you could call what we have an “IoT Cloud Platform.”
Wong: There seems to be a lot of development building blocks for the IoT available for the do-it-yourself (DIY) segment, including open source projects. How does this all compare with an IoT Cloud Platform?
Friedman: I think it’s really just a matter of project goals, and scale. If the goal is to connect a few things to the Internet, then tools such as Arduino or Raspberry Pi connected to a server are great. But when it comes to building products that scale to high volume, or that have complex data management requirements, it no longer becomes sufficient to cobble together building blocks.
An IoT platform such as Ayla’s, is built for ultra-high scalability, flexibility, and security. Moreover, platform providers put significant investment into testing and QA to make sure all of the pieces work together well. This, of course, results in a higher peace-of-mind for manufacturers launching (and need to support) products, and also a much faster time-to-market. It also means that executives, especially those at companies with significant brand value, can feel comfortable they are providing products that support the same quality they have built into their brand.
Wong: What some of the key considerations to evaluate in an IoT Cloud Platform?
Friedman: There are a couple major areas of focus that we see across our customers. One is flexibility: the customers don’t always know what features they want in their products, and they want the ability to very easily add/modify their features as they learn more from the market. Security is also a big one: this involves looking at all of the various “handshakes” that occur as devices communicate to the cloud and to applications, making sure there are no weak links in that exchange. Another consideration is scalability and uptime: customers want to feel comfortable that the service availability will meet their goals, and that the architecture uses all of the latest techniques to gracefully scale with high demand. Cost is also a driver, especially in the consumer markets. This involves the cost of the service as well as the total cost of ownership.
Finally, the overall completeness of the platform is important to customers. Manufacturers are increasingly seeking a platform that not only connects their devices to the Internet, but also helps them manage those devices, manage the data, connects to other cloud services, and provides them the ability to gain insights from the data.
This file type includes high resolution graphics and schematics when applicapable.
