Pervasive Intelligence: IoT Human-to-Machine and M2M Interaction Challenges

By Special Guest
Yen-Kuang (Y.K.) Chen, IEEE Fellow, IEEE IoT Volunteer, Panelist in the IEEE IoT Session at Upcoming IEEE Technology Time Machine Conference
September 27, 2016

Current trends point to the emergence of a fantastic future in less than a decade.

Advances in machine learning and artificial intelligence will enable pervasive intelligence in everyday devices to learn and dynamically support our preferences and lifestyles at home, at work and on the move.

The still somewhat vague phrase, the “Internet of Things” (IoT), captures this vision. Embedded processors already provide pervasive intelligence and we can expect increases of 4x-5x in current computational capabilities in 4-5 years due to Moore’s Law. Yet to achieve it, we still must meet many technical challenges and anticipate the ethical implications of technology’s impact on individuals and communities.

Today’s trends help frame tomorrow’s challenges.

Pervasive Intelligence
The Nest smart thermostat provides a good example of today’s technology. This device produces iterative responses that reflect a homeowner’s preferences for temperature settings. As those preferences change with the seasons, the Nest responds accordingly.

In similar fashion, the Nest Dropcam employs a digital video camera that might be set over one’s front door. Dropcam can be triggered by motion, but it can also go beyond motion detection to select the types of video content that triggers an alert to the homeowner.

Two devices, different roles, but each provide pervasive intelligence to humans.

Global Scale
Given the increase in computational abilities we expect from the embedded processors in everyday objects and the market for comfort and convenience, we can expect the number of devices within a home to proliferate. Home security, light and temperature settings, appliances and news and entertainment may all be enabled by distributed intelligence.

Soon, you and your family may interact with an autonomous network composed of several dozen points of pervasive intelligence, all within one household. Furthermore, some of its operational data must be coordinated with your neighbor’s home. And these homes operate within the context of the neighborhood, which resides in a town or city, which coordinates and collaborates with other cities.

Collaboration and Heterogeneity
What happens if the entire world – a global IoT – is connected in such a fashion? If devices were limited to just a smart thermostat and a security camera, little interaction among them would be needed. But we expect distributed intelligence to be embedded in many functional objects in a home, in a city, and beyond.

Thus we’re looking at a heterogeneous array of devices that will need to collaborate to ensure that the sum of their independent, autonomous actions equals the environment desired by the occupant. Potential conflicts or unintended consequences of disparate, autonomous devices acting in parallel will need resolution, either by the network itself or, less desirably, by human intervention.

Simply put, we will face challenges in coordinating networks of networks, each containing a heterogeneous array of autonomous devices designed to collaborate. The heterogeneity, collaboration, coordination and scale all pose daunting challenges that have yet to be resolved.

Perhaps the most glaring challenge is interoperability, enabled by global standards.

Security and Trust
Another top challenge would be designing and integrating mechanisms for security and trust, access and authentication, between human and machine and among machines in an autonomous network of networks. How much information about ourselves and our home or business premise and associated activities should we allow access to, and by whom?

Perhaps we’d like our friends and family to know when we’re home. But it’s just as likely that we might want that information blocked to some friends and family. What about access by third parties? If you’re disturbed by seeing advertisements on your search engine results page for products you recently researched, think of the barrage of unwanted attention third parties could produce if they can see the status of your appliances and reach you with suggestions.

Back to the Vision
These are difficult challenges for the best minds. But the market incentives involved also exist on an unprecedented scale. I’d argue that every effort will be expended to meet these challenges.  

If we can solve the issues presented by large-scale heterogenenous collaboration and security/trust, our collective future may look very different in just five to 10 years. 

Edited by Ken Briodagh
Get stories like this delivered straight to your inbox. [Free eNews Subscription]

Related Articles

Beyond the Closet, Connecting to IoT

By: Gary Audin    11/11/2020

Two challenges arise when considering cable based IoT.

Read More

Banyan Security Enhances Secure Remote Access for Engineering Resources

By: Ken Briodagh    10/27/2020

Banyan's Continuous Authorization Can Grant or Revoke Access to Sensitive Engineering Environments and Applications in Real-time Based on TrustScore

Read More

Senet Eyes RAN Partnerships as Key to Delivering Network Services for Massive IoT

By: Arti Loftus    10/21/2020

To meet the challenges that come with providing network connectivity for IoT solutions, Senet is executing a strategy for massive IoT that will be bui…

Read More

mimik Selected by 5G Open Innovation Lab to Drive Early Adoption of 5G

By: Ken Briodagh    10/15/2020

mimik's patented Hybrid Edge Cloud platform will boost the performance and reduce the cost of 5G Networks

Read More

5G Sets New Standards for Vertical Industries' IoT Connectivity

By: Special Guest    10/13/2020

As 5G rolls out across the world, vertical industries across IoT are working on additional standards to make the technology suitable for their industr…

Read More