<img alt="" src="https://secure.ruth8badb.com/159331.png" style="display:none;">
Cimpl Blog

Cimpl Blog

Your Weekly Insights on How to Manage your Enterprise Digital Footprint.

Smartphones become Superphones – The Next Step in Workplace Evolution

Posted by Henry Cheang | June 11, 2014 8:00 AM
Scientists have been trying to get robots to think and act and see like humans…

For years now, researchers have been working to develop technology that can recognize objects and events in the same way that humans do. There are software and robots that navigate their own path across different terrains, map out sound sources, and which have actually been programmed to be curious. This research has been, by and large, pretty successful.

Now, how does this impact us, IT and business professionals? Well, we’re about to see smartphones become superphones. There’s a fantastic new microchip – the nn-X chip – that can, in real-time, identify and distinguish between different images, videos, and objects. Including people and faces. I think you can see where I’m going with all this, right? I’m going to dial this back a bit before going any further, so that we can get a bit of background first.

What is the nn-X chip?

Eugenio Culurciello, a professor at Purdue University, created the nn-X chip. When this chip is given input from a camera, everything in the camera’s field of view is immediately (for human intents and purposes) identified and tagged. This classification of the image allows the scene to be thoroughly analysed and contextualized in real time.

This method is called “deep learning” because it processes information via multiple layers of neural networks (learning algorithms). Effectively, it’s a system of machine learning that attempts to mimic the way the human brain analyzes information. To take the example of facial recognition, these are the steps the method would take to identify a human face:

  • An image is inputted to the deep learning system
  • The system separates the image into layers and reduces image resolution for easier computation
  • Feature extraction is done by preprogrammed algorithms
    • Separate algorithms identify different elements per layer, e.g., one layer recognizes eyes, another the nose, and so on. The identifications are combined.
  • The process is continued until the face is recognized

The overall process isn’t really that new - internet companies have been using deep-learning software for some time already as a way for its users to search the web for media (i.e., pictures and video) which have been tagged with keywords. Unfortunately, deep learning wasn’t possible on small-scale devices due to computation restrictions. Until now.

Culurciello and his team have so thoroughly refined the nn-X chip’s layer-processing functions that the chip operates at several tens of times faster than those found on laptop chips, and consumes 10 times less power than that of commercial processors. And this is what makes it possible to integrate deep-learning into smartphones (and making them superphones). According to Culurciello, superphones (i.e., in this case, smartphones embedded with the nn-X chip) can be useable next year.

WWSD (What would a superphone do)?

Well… Just the facial recognition example above would do wonders. Basically, once you’ve identified someone, it becomes pretty trivial to get all publicly-available information on them. In this day and age, that would typically be a lot of information.

Imagine that you’re the new guy (or gal) at work in a company. Wouldn’t it be great to ascertain common points of professional strengths or personal interests on your first day? Deep-learning should be able to do that – it would identify your new coworkers by face, and then use that information to determine her or his interests through the information s/he has posted on social media. With that in hand, you’d have some easy topics with which to break the ice.

On a deeper level, imagine you’re at an industry event or a business lunch. An nn-X chip-enabled phone could help you detect and prioritize everyone in the room who’s a prospective lead, based on her or his identity and online presence.

And those were the easy examples. Deep learning technology does much more than simple facial recognition – it can identify all manner of details. By doing the legwork of automatically identifying the things in your environment, it gives you a head start on better understanding the context in which you find yourself. Ideally, nn-X chip-enabled phones should serve as your own automated research clerk. Wouldn’t that just be super?

Back to the present…

Now, unfortunately, we don’t quite have this technology in hand just yet. What we do have, however, is the power of business process automation (BPA). BPA has already been changing the face of organizations, big and small. And we at Cimpl really understand technology and automation – we pride ourselves on being  Canada’s leader in technology expense management, after all. Contact us today to find out more about technology management!


New Call-to-action



Topics: Mobile Devices, Technology Expense Management (TEM), Automation, Smartphones

Written by Henry Cheang

Henry is a dedicated technical writer, focused on conducting market research, contributing to product design, and writing clear and concise documentation for the company. He is an enthusiastic team member and is passionate about science and technology, who plays a key role in Cimpl’s product messaging. His dedication to writing is reflected in his experience in authoring academic papers, documentation, user guides, and in contributing to Cimpl’s marketing efforts.

Subscribe to our Blog

Follow Us!

2019 Product Roadmap Webinar
What is WEM? ePaper
What is UCCEM?