Posts Tagged ‘connectivity’

Will Moore’s and Metcalfe’s Laws Cross the IOT Chasm?

Sunday, April 30th, 2017

The success of the IOT may depend more on a viable customer experience over the convergence of the semiconductor and communication worlds.

By John Blyler, Editor, IOT Embedded Systems

The Internet of Things will involve a huge number of embedded devices reporting back to data aggregators running servers on the cloud. Low cost and low power sensors, cameras and other sources will allow the IOT to render the real world into a digital format. All of these “things” will be connected together via the Internet, which will open up new business models and services for customers and users. It should greatly expand the human–machine experience.

The key differentiators between the emerging IOT and traditional embedded systems is connectivity. IOT will conceivable connect all embedded things together. The result will be an almost inconceivable amount of data from sensors, cameras and the like, which will be transferred to the cloud for major computation and analysis.

Connectivity means IOT platforms will have a huge data side. Experts predict that the big data industry will grow to about US$54.3 billion by 2017. But the dark side of connectivity is the proliferation of hacking and privacy lapses caused by poor security.

Security is an issue for users as well as for the device developers. Since most IoT devices are resource constrained, designer cannot deploy resource-intensive security protection mechanisms. They are further constrained by the low cost of mass-produced devices.

Another challenge is that most software developers are not particularly security or reliability conscious. They lack training in the use of security testing, encryption, etc. Their code is often not design nor programmed in a defensive fashion.

Finally, since IOT devices will be designed and available on a massive scale, security attacks and failures can be easily propagated. Frequently software security patches will be needed but these must be design for early in the development life cycle of both the hardware (think architecture and memory) and software.

Moore-Metcalf and the Chasm

Connectivity, security and data analysis will make IOT devices far more complex than tradition embedded systems. This complexity in design and product acceptance can be illustrated by the confluence of two laws and a marketing chasm. Let’s consider each separately.

First, there is Moore’s Law. In 1965, Intel co-founder Gordon Moore predicted that transistor density (related to performance) of microprocessors would double every 2 years (see Figure 1). While “doubling every 2 years” suggests a parabola-shaped curve, Moore’s growth function is almost always represented in a straight line ― complemented by a logarithmic scale on the Y-axis.

Figure 1: Moore’s Law (courtesy of Mentor Graphics, Semi Pacific NW, 2015)

Several years later, another technology pioneer, 3Com co-founder Bob Metcalfe, stated that the value of a network grows with the square of the number of network nodes (or devices, or applications, or users, etc.), while the costs follow a more or less linear function. Not surprisingly, this equation is show as a network connection diagram. For example, 2 mobile devices will only able to communicate with each other. However, if you have billions of connected devices and applications, connection complexity rising considerably (see Figure 2).

Figure 2: Metcalfe’s Law.

Metcalfe’s Law is really about network growth rather than about technological innovation. Blogger Marc Jadoul recently noted on the Nokia website that, the combination of Moore’s and Metcalfe’s principles explains the evolution of communication networks and services, as well as the rise of the Internet of Things. The current IoT growth is enabled by hardware miniaturization, decreasing sensor costs, and ubiquitous wireless access capabilities that are empowering an explosive number of smart devices and applications…”

Jadoul realizes that the availability of state-of-the-art technology does not always guarantee success, citing the struggling growth of two main IOT “killer” consumer devices and apps, namely, watches and connected thermostats. The latter is also notorious for its security issues.

He explains this slow adoption by considering the “chasm.” Geoffrey A. Moore wrote about the gap that product marketers have to bridge for a new technology to go mainstream. Jadoul then combines these three charts, admitting the inaccuracies caused by different axis and scales, to observe that the chasm is actually the point where the shift from a technology driven model to a value and customer experience driven business needs to take place (see Figure 3).

Figure 3: Intersection of Gordon Moore’s Law, Metcalfe’s Law and Geoffrey Moore’s “the Chasm. (Courtesy of Marc Jadoul blog.)

This line of reasoning highlights the key differentiator of the IOT, i.e., connectivity of embedded semiconductor devices. But the success of the IOT may depend more on a viable customer experience over the convergence of computational and communication technologies.

Cybernetic Human Via Wearable IOT

Tuesday, January 17th, 2017

UC Berkeley’s Dr. Rabaey sees humans becoming an extension of the wearable IoT via neuron connectivity at recent IEEE IMS event.

by Hamilton Carter and John Blyler, Editors, JB Systems

During the third week in May, more than 3000 microwave engineers from across the globe descended upon San Francisco for the International Microwave Symposium 2016. To close the week, it seemed only fitting then that the final plenary talk by Jan Rabaey was titled “The Human Intranet- Where Swarms and Humans Meet.”

RabaeyImg_rotate-crop

Dr. Rabaey, Professor and EE Division Chair at UC Berkeley, took the stage wearing a black T-shirt, a pair of slacks, and a sports coat that shimmered under the bright stage lights. He briefly summarized the topic of his talk, as well as his research goal: turning humans themselves into the next extension of the IoT. Ultimately he hopes to be able to create human-machine interfaces that could ideally not only read individual neurons, but write them as well.

What Makes a Wearable Wearable?

The talk opened with a brief discourse on the inability thus far of wearables to capture the public’s imagination. Dr. Rabaey cited several key problems facing the technology: battery life; how wearable a device actually is; limited functionality; inability to hold user interest; and perhaps most importantly something he termed stove-piping. Wearable technologies today are built to communicate only with other devices manufactured by the same company. Dr. Rabaey called for an open wearables platform to enable the industry to expand at an increasing rate.

Departing from wearables to discuss an internet technology that almost everyone does use, Dr. Rabaey focused for a few moments on the smart phone. He emphasized that while the devices are useful, the bandwidth of the communications channel between the device, and its human owner is debilitatingly narrow. His proposal for remedying this issue is not to further enhance the smart phone, but instead to enhance the human user!

One way to enhance the bandwidth between device and user is simply to provide more input channels. Rabaey discussed one project, already in the works, that utilizes Braille-like technology to turn skin into a tactile interface, and another project for the visually-impaired that aims to transmit visual images to the brain over aural channels via sonification.

Human limbs as prosthetics

As another powerful example of what has already been achieved in human extensibility, Dr. Rabaey, showed a video produced by the scientific journal “Nature” portraying research that has enabled quadriplegic Ian Burkhart to regain control of the muscles in his arms and hands. The video showed Mr. Burkhart playing Guitar Hero, and gripping other objects with his own hands; hands that he lost the use of five years ago. The system that enables his motor control utilizes a sensor to scan the neurons firing in his brain as researchers show him images of a hand closing around various objects. After a period of training and offline data analysis, a bank of computers learns to associate his neural patterns with his desire to close his hand. Finally, sensing the motions he would like to make, the computers fire electro-constricting arm bands that cause the correct muscles in his arm to flex and close his hand around an object. (See video: “The nerve bypass: how to move a paralysed hand“)

Human Enhancements Inside and Out

Rabaey divides human-enhancing tech into two categories, extrospective, applications, like those described above, that interface the enhanced human to the outside world, and introspective applications that look inwards to provide more information about enhanced humans themselves. Turning his focus to introspective applications, Rabaey presented several examples of existing bio-sensor technology including printed blood oximetry sensors, wound healing bandages, and thin-film EEGs. He then described the technology that will enable his vision of the human intranet: neural dust.

The Human Intranet

In 1997, Kris Pister outlined his vision for something called smart dust, one cubic millimeter devices that contained sensors, a processor, and networked communications. Pister’s vision was recently realized by the Michigan Micro Mote research team. Rabaey’s, proposed neural dust would take this technology a step further providing smart dust systems that measure a mere 10 to 100 microns on a side. At these dimensions, the devices could travel within the human blood stream. Dr. Rabaey described his proposed human intranet as consisting of a network fabric of neural dust particles that communicate with one or more wearable network hubs. The headband/bracelet/necklace-borne hub devices would handle the more heavy-duty communication, and processing tasks of the system, while the neural dust would provide real-time data measured on-site from within the body. The key challenge to enabling neural dust at this point lies in determining a communications channel that can deliver the data from inside the human body at real-time speeds while consuming very little power, (think picowatts).

Caution for the future

In closing, Dr. Jan implored the audience, that in all human/computer interface devices, security must be considered at the onset, and throughout the development cycle. He pointed out that internal defibrillators with wireless controls can be hacked, and therefore, could be used to kill a human who uses one. While this fortunately has never occurred, he emphasized that since the possibility exists it is key to encrypt every packet of information related to the human body. While encryption might be power-hungry in software, he stated that encryption algorithms build into ASICs could be performed at a fraction of the power cost. As for passwords, there are any number of unique biometric indicators that can be used. Among these are voice, and heart-rate. The danger for these bio-metrics, however, is that once they can be cloned, or imitated, the hacker has access to a treasure-trove of information, and possibly control. Perhaps the most promising biometric at present is a scan of neurons via EEG or other technology so that as the user thinks of a new password, the machine interface can pick it up instantly, and incorporates it into new transmissions.

Wrapping up his exciting vision of a bright cybernetic future, Rabaey grounded the audience with a quote made by Joanna Zylinska, an Australian performance artist, in a 2002 interview:

“The body has always been a prosthetic body. Ever since we developed as humanoids and developed bipedal locomotion, two limbs became manipulators. We have become creatures that construct tools, artifacts, and machines. We’ve always been augmented by our instruments, our technologies. Technology is what constructs our humanity. …, so to consider technology as a kind of alien other that happens upon us at the end of the millennium is rather simplistic.”

The more things change, the more they stay the same.

Clarifying Embedded IOT Connectivity Confusion

Tuesday, June 28th, 2016
YouTube Preview Image

Clarifying Embedded IOT Connectivity Confusion


Extension Media websites place cookies on your device to give you the best user experience. By using our websites, you agree to placement of these cookies and to our Privacy Policy. Please click here to accept.