Ambient computing is a broad term that describes an environment of smart devices, data, AI decisions, and human activity that enables computer actions alongside everyday life, without the need for direct human commands or intervention. Ambient computing represents an unparalleled opportunity to enhance almost every sphere of society – from the professional to the personal. And in my opinion, it is also the ultimate use case for which semantic knowledge graphs were created.
With knowledge graph standards, ambient computing is no longer a mere ideal or science fiction fantasy on television or in books. It’s a real computational model involving Internet of Things (IoT) endpoints, AI analytics, machine reasoning, orchestration, and low latent event processing at the edge to anticipate users’ desires and perform timely action – without explicit commands.
JOIN US FOR THIS LIVE ONLINE DATA MODELING COURSE
Want to learn the fundamental building blocks of Data Modeling? Take a look at this three-day training.
For example, a motion detector might identify a homeowner’s return from work at night, open the garage accordingly, and trigger a thermostat to increase the air conditioning to a desired temperature while smart gadgets in the kitchen begin preheating the oven for dinner.
Each of these actions happens without someone deliberately engaging with these disparate systems. One’s interactions with his or her environment dictate which events occur, relegating the computational process to the background to benefit humans.
Different vendors currently have varying degrees of ambient computing in place. Amazon has several household devices that interact with Alexa, for example. Still, the larger vision of ambient computing can’t be restricted to one vendor and must include timely data exchanges between vendors, products, and operating systems.
Doing so requires systemic interoperability, the likes of which the universal standards powering semantic graph technology have provided for years. This smart data approach is integral to the mainstream adoption of ambient computing, which is impending.
Data Management Requirements
Ambient computing involves several key facets of Data Management, foremost of which is the IoT’s inter-device connectivity, various sensors and actuators for devices, and connections with central cloud locations. The IoT’s real-time responses are based on a tandem of cognitive computing analytics and machine reasoning that anticipates users’ needs before fulfilling them.
There’s also a vital data integration layer for the assortment of technologies generating different data types, some of which may be proprietary. Finally, run-time orchestration capabilities (partially dependent on active Metadata Management) are necessary to formulate the action required from the rapid event processing taking place in these IoT endpoints or the cloud. Each of these functions, particularly the inter-device communication at the edge, requires uniform standards for interoperability.
Interoperable Standards
Knowledge graph technology offers the most capable standards for meeting these demands. By describing data with semantic statements and ascribing Uniform Resource Identifiers (URIs) to each node or datum, this approach harmonizes data of all types while making them inherently machine-readable – which is ideal for ambient computing. This standards-based approach is also fortified by reasoning capabilities in which the underlying systems make intelligent inferences about their data.
Such functionality can spur integration efforts by automatically addressing differences in schema between manufacturers, for example, so their respective devices can effectively communicate. It’s also essential for implementing rules so that, for example, if an elderly person with a wearable device falls, alerts are sent to family members’ smartphones as well as to health care personnel monitoring his or her vital signs to contact appropriate response personnel.
Uniform Semantics
The most important aspect of the semantic standards at the core of knowledge graph technologies is their unification of the diverse business concepts, terminologies, and definitions between all types of data for much-needed agreement about the meaning of specific data between systems. There are several uniform semantic data models – which naturally evolve to include additional sources, terms, and business requirements – that provide this functionality. These ontologies involve a hierarchical understanding of the specific words represented in data, resulting in uniform taxonomies.
This knowledge is necessary so devices of respective manufacturers and areas of interest (like a video surveillance camera in an organization’s office and smart locks or smart lighting in a smart building) can share data about an intruder to dynamically secure valuable areas, for example. Most significantly, knowledge graphs can align data of all variations according to these standardized data models and taxonomies to connect data for optimal understanding of how even the most diverse datasets interrelate for ambient computing use cases. Examples might include autonomous vehicles receiving real-time updates about traffic patterns in smart cities to adjust their routes to deliver goods more efficiently or deposit passengers at the airport in time to make their flights.
The Ambient Reality
Ambient computing will never become a reality until universal standards are adopted for semantic interoperability. Knowledge graphs have incorporated such standards for machine intelligence since their inception. As such, they’re uniquely positioned to make the goal of intuitive computing based on a user’s fluid interactions with his or her environment a reality, which simply won’t be possible without these standards.
With critical mass achieved for standards, the next question is, who is actually going to do the “ambient computing” for the end user? We may see one of the big three (Google’s assistant, Apple’s siri, Amazon’s Alexa, or maybe even Samsung) or a slew of new companies come up that will try to own this space. It is also possible that we’ll put control in the hands of the end users, perhaps via Tim Berners-Lee’s Solid stack, which is open source and completely based on semantic web standards, giving it a head start.
