Menu

IoT FEATURE NEWS

Down to the Device, Data Analytics as a Service AI as the Real Breakthrough

By

With the mass and moving parts that will define the IoT, can human beings really keep up? Or is it time to let the machines do the thinking?

There is no point in connecting things unless those things send information that help those who own and operate those things optimize their value.

Consider the difference between an office building before and after a full IoT installation. A legacy office building can operate with mechanical features: keys to unlock doors, switches to turn lights off and on, a network connecting phones and computers, a refrigerator and coffee maker in the kitchen. Multiply this by a dozen tenants, and this legacy office building can be rented and secured by the landlord just as that landlord and landlords before did so.

But a smart office building changes the experience and economics for the businesses who rent space, the property management company responsible for keeping the spaces filled, and the facilities management company responsible for ensuring operations and security, while also meeting quality of service and budget goals.

Here’s a short list of what can be instrumented in a smart building with a comprehensive IoT approach:

  1. Main door security
  2. Tenant door security
  3. Private office security
  4. Conference room security
  5. Surveillance cameras
  6. Biometric access nodes
  7. HVAC system management and control
  8. Lighting management and control
  9. Information kiosks
  10. Visitor office usage
  11. Employee office usage
  12. Conference room usage
  13. Conference room scheduling
  14. Video conferencing equipment and service availability
  15. WiFi network availability and capacity management
  16. Bathroom facility water usage
  17. Leak detection
  18. Refrigerator temperature control
  19. Coffee and tea station inventory and control
  20. People counting
  21. Computers and other equipment tracking and management
  22. Windows open/closed
  23. Blinds open/closed
  24. PA system
  25. Emergency response system

Multiply all these services across multiple locations, and you’ll soon figure out that making the smarter office even smart with AI makes sense.

Shamshad (Sam) Ansari, co-founder and CEO of Accure Analytics, looks at nearly everything from the perspective of AI when it comes to the IoT and especially the IIoT.

“To realize the full potential of IoT’s value, enterprises are combining IoT with AI technologies, which enable connected things to simulate intelligent behaviors and act, increasingly with little or no human intervention,” Ansari said. “Why spend tens or hundreds or even millions of dollars to instrument an environment, like a smart office, if you cannot maintain and control all those endpoints efficiently?”

Integrating AI into IoT networks is not new, but neither is it pervasive, as IoT itself is being continually redefined and refined. “Every IoT company out there must move rapidly to identify how they’ll drive value from combining AI and IoT, or they will find themselves extinct,” Ansari said, “and the same can be said for CIOs and others in enterprises who are undertaking IoT projects.

The personal computer was cool when it first came on the scene, but it wasn’t until software applications and connectivity to the Internet ramped up that the PC disrupted everything and forced businesses into the dawning of the digital era.

The same can be said for the IoT – only this time there are exponentially more endpoints and functions, and exponentially more innovation coming out of what we know is now digital enterprises on steroids.

“The profound impact that AI will have on every aspect of our personal and working lives is magnified when it is combined with sensors that allow things to communicate, and it is the coming era of real time IoT that making AI and machine learning absolutely critical, rather than an added feature,” Ansari said.

AI is forecast as a massive market; for example, Bank of America Merrill Lynch (BAML) estimated in December 2015 that the robot and AI solutions market will surge to $153 billion a year by 2020, with $70 billion of that spent on AI-based analytics. A primary driver? Cost savings.

Cost savings due to needing fewer humans to manage things, particularly the most mundane things that AI systems, for example in smart buildings, can manage very precisely, reliably and automatically.

“We’ve moved well beyond the early stages of AI and are now in the era of active intelligence—the ability to act in real time with little or no human intervention,” Ansari said. “For the IoT, scaling active intelligence makes solutions very attractive, with almost immediate returns on investments particularly when it comes to smart cities, smart buildings and smart homes. Given the power and scalability of a proper AI solution within an IoT deployment, tasks that used to take humans weeks or months to complete can be done continuously and in seconds.”

Accure was one of six companies chosen by the Smart City Works non-profit organization, which focuses on smart infrastructure that will make urban cities more livable and environmentally sustainable. “We liked the direction Accure was heading in as we found value in embedding AI into nearly every solution we studied while preparing for the first consort of companies as part of our program,” said Bob Mazer, co-founder of Smart City Works.

In fact, Mazer pointed out that just as the world realized “mobile first” made sense ten years ago when mobile enterprise communications and computing started to surpass desktop innovation, “AI first” is becoming a key strategy for enterprises undertaking large scale smart building and smart factory implementations, and governments and organizations are building increasingly more sophisticated smart city systems.

“It’s logical – the more things we add sensors to, the more there is to manage, so automating as we build is the only sustainable way to ensure IoT and IIoT investments realize the cost savings or improvement in experience they envision in the conceptual phases,” Mazer continued.

As enterprises and government agencies accelerate their embrace of IoT, it becomes increasingly inevitable that AI must play a pivotal role. IoT is primarily about data, data is the currency of the digital economy and the IoT is generating and will continue to expand growth in data volumes, another important challenge for AI providers to tackle.

“Data is only useful if it is actionable,” Ansari said, “and to be actionable, there must be contextual awareness, and a means to set policy and govern how the system can become as automated as possible, while still providing humans information they can manage their lives or businesses better with.”

Ansari  pointed out the demand for software and analytics which enable real time responses, for example video cameras reading license plates or shapes of cars for parking lot compliance (reducing lost revenue due to lost ticket schemes) and secondly in post-event processing, seeking out patterns in data over time and running predictive analytics for maintenance and other requirements.

“Predictive analytics asks what will happen, prescriptive analytics drives certain actions, and continuous analytics enables the machine learning that enables IoT systems to run better and better over time,” Ansari explained. “IoT also helps enrich existing AI systems, as sensors provide real time input and feedback. The interdependence between IoT and AI also works the other way. IoT’s capacity to enable real-time feedback; engineered properly, this symbiotic relationship can generate enormous value, including reducing opex by enabling a single person, for example, to now manage a dozen smart buildings from a centralized view.”

Accure has been rolling out smart IoT solutions using its Momentum platform, which Ansari says is heavily weighted toward security and scalability. “No IoT AI software will succeed unless the architecture provides for massive scalability – millions of concurrent devices – along with batch analytics for machine learning model development. We’re seeing increasing interest in real time streaming analytics as a subscription service, and useful features like sending alerts and notifications, and feeding business applications, particularly in the manufacturing realm. Smart maintenance alone can save some of the enterprises we are working with millions of dollars a year once the systems are brought up across the global enterprise.”

While Ansari could not reveal specifics, one project underway is predicting oil pump failures sixteen hours in advance, which can literally save millions of dollars due to unnecessary downtime. Using three years of historical data, the algorithms were written and proven to work, “and the system will only get better as more and more machine learning takes place.”

Accure has also implemented smart parking with sensors combined with meters and mobile apps, all devices collecting data that fed into an AI system enabling a municipality to better manage parking availability while also improving compliance at a lower cost (compared to “meter readers.”)

Safety measures have also driven business for Accure, including in airports and parking garages.

When asked how their data scientists are creating a competitive offering compared to the more well-known engines, like IBM Watson, Ansari said “IBM Watson is a pure cloud-based, self-service system; in order to build a real solution on Watson you need to hire data scientists. Our platform includes the algorithms – no data scientist required – just a solid business analyst who can work from our console which is very intuitive and attractive to IT staff.”

Secondly, Ansari says many enterprises are still hesitant to send data to a shared cloud, like IBM Watson. “Our solution can work on any cloud – public, private, hybrid – or on prem – we are neutral and do whatever makes sense for our customers. We also can help them build proprietary AI applications with them, and in fact work with IoT sensor companies enabling them to offer end-to-end smart solutions with AI embedded.”

Ansari does make one think that IoT really is the next version of the PC jump shift for enterprises, particularly when software applications and practical, cost efficient features are built in – including Artificial and Actionable Intelligence.




Edited by Ken Briodagh

Contributing Writer

SHARE THIS ARTICLE
Related Articles

MachineMetrics Announces $20M Series B Funding Round

By: Arti Loftus    6/17/2021

The growth of applications designed to further automate and optimize manufacturing operations continues unabated given the success of modernization pr…

Read More

New Statistics from The Eclipse Foundation Indicate that Edge Computing Adoption Continues to Boom

By: Matthew Vulpis    6/14/2021

The Eclipse Foundation, one of the world's largest open-source foundations, recently announced the availability of its 2021 IoT and Edge Commercial Ad…

Read More

An Edge Computing Breakup: Out with the Old, and In with the New

By: Special Guest    6/9/2021

When COVID-19 arrived in early 2020, enterprises' first priority was to patch together a communications and information-sharing infrastructure that co…

Read More

Cloud-Based Cellular Network Platform Challenger Monogoto Lands $11M in Funding

By: Arti Loftus    6/9/2021

Today, Monogoto, innovator, and developer of a secure, global cloud-based cellular network platform, announced they have closed a round including vent…

Read More

As Linux Foundation's Zephyr Project Turns Five, Addressing Constrained Device Challenges is More Important Than Ever

By: Arti Loftus    6/7/2021

Noting nearly 1,000 contributors, 50,000 commits building advanced support for multiple architectures including ARC, Arm, Intel, Nios, RISC-V, SPARC a…

Read More