According to IoT Analytics, there are over 17 Billion connected devices in the world as of 2018, with over 7 Billion of these “internet of things” (IoT) devices. The Internet of Things is the collection of those various sensors, devices, and other technologies that aren’t meant to directly interact with consumers, like phones or computers. Rather, IoT devices help provide information, control, and analytics to connect a world of hardware devices to each other and the greater internet. With the advent of cheap sensors and low-cost connectivity, IoT devices are proliferating.
It is no wonder that companies are inundated with data that comes from these devices and are looking to AI to help manage the devices as well as gain more insight and intelligence from the data exuded by these masses of chatty systems. However, it is much more difficult to manage and extract valuable information from these systems than we might expect. There are many aspects and subcomponents to IoT such as connectivity, security, data storage, system integration, device hardware, application development, and even networks and processes which are ever changing in this space. Another layer of complication with IoT has to do with scale of functionality. Often times, it’s easy to build sensors to be accessed from a smart device but to create devices that are reliable, remotely controlled and upgraded, secure, and cost effective is a much more complicated matter.
How AI is Transforming IoT
On a recent AI Today podcast, Rashmi Misra from Microsoft shared how AI and IoT are combining to provide greater visibility and control of the wide array of devices and sensors connected to the internet. At Microsoft, Rashmi leads a team that builds IoT and artificial intelligence (AI) solutions, where she works across partners of all sorts such as device manufacturers, application developers, systems integrators, and other vertically focused partners who want to play key AI technologies in IoT fields. Her Microsoft team is focused on gaining insights and knowledge from data that is created from IoT devices, simplifying the access and reporting of that data. (Disclosure: I am a host of the AI Today podcast)
IoT is transforming business models by helping companies move from simply making products and services to companies that give their customers desired outcomes. By impacting organizations’ business models, the combination of IoT-enabled devices and sensors with machine learning creates a collaborative and interconnected world that aligns itself around outcomes and innovation. This combination of IoT and AI is changing many industries and the relationships that businesses have with its customers. Businesses can now collect and transform data into usable and valuable information with IoT.
As an organization applies digital transformation principles to its business, the combination of IoT and AI can create a disruption within its industry. Whether an organization is using IoT and AI to engage customers, implement conversational agents for customers, customize user experiences, obtain analytics, or optimize productivity with insights and predictions, the use of IoT and AI creates a dynamic where companies are able to gain high quality insight into every piece of data, from what customers are actually looking at and touching to how employees, suppliers, and partners are interacting with different aspects of the ecosystem. Instead of just having the business processes modeled in software in a way that approximates the real world, IoT devices give systems an actual interface to the real world. Any place where you can put a sensor or a device to measure, interact, or analyze something, you can put an IoT device connected to the AI-enabled cloud to add significant amounts of value.
Using AI to Help Make Sense of IoT Data
Common challenges organizations face today with AI and IoT are with application, accessibility, and analysis of IoT data. If you have a pool of data from various sources you can run some statistical analysis with that data. But, if you want to be proactive in predicting events to to take future actions accordingly such as when to change a drill bit or anticipate a breakdown in a piece of machinery, a business needs to learn how to use these technologies to apply them to discern this kind of data and process.
The sheer quantity of IoT data, especially in organizations that have deployed sensors or tags down to the individual unit level is significant. The massive amount of constantly changing data is too difficult to manage with traditional business intelligence and analytics tools. This is where AI steps in. Through the use of unsupervised learning and clustering approaches, machine learning systems can automatically identify normal and abnormal patterns in data and alert when things deviate from observed norms, without requiring advance setup by human operators. Likewise, these AI-enabled IoT systems can automatically surface relevant insights that might not be visible for the haystack of data that makes those insights almost invisible.
Enterprises are implementing AI-enabled IoT systems in a number of different ways. Solutions firms are producing prepackaged code and templates that include tried and tested models for specific application domains such as shipping and logistics, manufacturing, energy, environmental, building and facilities operations, and other models. Others are creating customer solutions building and training their own models, leveraging cloud-providers to harness external CPU power. Some solutions centralize AI capabilities in on-premise solutions or cloud-based offerings, while others aim to decentralize AI capability, pushing machine learning models to the edge to keep the data close to the device and speed up performance. There are a number of ways to implement this technology and the challenge is in applying it and appropriately accessing it.
Today, we are seeing a lot of growth with both AI and IoT. These technologies combine to enable the next level of automation and productivity while decreasing costs. As consumers, businesses, and governments start to control IoT in a variety of environments our world will change greatly and allow us all to make better choices. It’s already rapidly changing everything from retail to supply chain to health care. AI-enabled IoT is transforming the energy industry with smart energy solutions, where a city or town wants to create a delocalized power trade due to houses with solar panels. Rashmi shares an example of how IoT is changing supply chain and logistics. In that example, milk is very susceptible to changes in temperature. If you produce the milk and it gets transported from one place to another, you can use IoT to track the humidity of the environment during the transportation of the milk every step of the way. The private and public sector stand to have a huge impact by gaining more intelligence from all the devices out there.
The proliferation of IoT devices is making the future a very connected and instant access to information world. There is now need for AI to manage all those devices and make sense of the data that comes back from them. In these ways, AI and IoT are very symbiotic and will continue to have an intertwined relationship moving forward.