There is innovation in convergence. Look at the convergence of hardware and software, electrical and mechanical engineering, or data and design. As multiple ideas and technologies come together, they can create something greater than the sum of their parts.
That’s the exciting side of designing products for the Internet of Things (IoT): making almost any product suddenly “smart” and innovative via the convergence of ubiquitous sensing and communication. It’s easy to imagine the possibilities, such as cars that sense where people are and refuse to hit them, parts that sense when they’re about to break and call in a service technician, or products that tell their design engineers how they’re actually being used out in the field so that he/she can improve upon them. The possibilities are so mind boggling that it’s easy to forget the significant challenges involved.
The flip side of convergence is complexity. More connected “things” require more lines of code, more bandwidth for communication, and produce more data to collect and analyze. Much of that complexity is often shouldered by the engineers who are responsible for bringing new technologies to bear, collaborating among converging disciplines and departments, and making use of vast amounts of data — all in shorter and shorter design cycles. Who will overcome the challenges involved in making the IoT a reality? According to National Instruments’ NIWeek conference theme, “You and NI Will.”
The 20th annual NIWeek that took place this week in Austin, TX, was attended by more than 3,000 engineers and researchers. Over four days, the company announced LabVIEW 2014, the latest incarnation of its graphical system design software; the new CompactRIO software-designed controller, which runs on an Intel Atom processor and Xilinx Kintex-7 FPGA; the PXI-based Semiconductor Test System; and a new CompactDAQ 4-slot controller, also equipped with an Intel Atom processor; among others. All of the products were presented as a means to decrease complexity in the design process so that engineers and scientists could meet the grand engineering challenges.
Timing is Everything
The question of “when” came up during NI co-founder Dr. Jim Trouchard’s opening keynote. Not the when of making the IoT a reality, but the when of tight timing. Much of the IoT relies on timing, which is not standard among devices and machines that need to communicate with one another. Effective measurement and control depends on timing, the higher resolution the better.
Dr. T, as he’s known in the industry, said NI’s focus on timing and synchronization has given it a distinct advantage in the marketplace.
Capturing real-time, real-world data, what NI calls Big Analog Data, is just part of the challenge. Turning that data into “wisdom,” as DR. T. called it, often requires integration of hardware timing at the sub-nanosecond level, global timing and FPGA timing, he says.
Dr. T: NI is making investments in 2 key areas to differentiate its platform: 1. Distributed time 2. Multirate/parallel Computation. #NIWeek
— Desktop Engineering (@DEeditor) August 5, 2014
To help keep humans in the real-time data loop, NI will be releasing NI InsightCM Enterprise in October. It will allow engineers to collect, share and collaborate on conditioning monitoring data as it’s generated. It can “acquire and analyze sensor data, generate alarms and allow specialists to diagnose results and draw conclusions,” said Kamalina Srikant, NI’s product marketing manager, Embedded Systems, while introducing NI InsightCM Enterprise at NIWeek.
The company has also differentiated itself with timing on the software side with LabVIEW, Dr. T said. “Traditional programming languages do not deal with time,” he said. “LabVIEW does.” He noted LabVIEW’s time loop that allows it to manage multicore systems and distribute processing between the processor and FPGA. For example, hybrid electric vehicle simulation requires high-speed computations that are synchronized with the rest of the simulation to enable much faster development time. You can watch Dr. T’s presentation in its entirety here:
To see why and how Subaru used NI LabVIEW to speed the testing of its XV Crosstrek Hybrid vehicle, watch the brief video below. The project was a Grand Challenges Award winner in the 2014 NI Engineering Impact Awards. Read more about it here.
Wireless latency also affects timing. As the IoT drives ubiquitous connectivity, the speed with which devices are able to communicate becomes increasingly important. If you’ve ever been on a conference call with a few seconds of lag between when the speaker talks and you hear him, you know firsthand how that affects communication. Imagine the effect of that delay on a car that relies on receiving a signal from a pedestrian’s cell phone to know that person is crossing the street in front of the vehicle.
Making speed even more challenging to deliver is the fact that mobile traffic has been close to doubling annually. The mobile industry predicts mobile traffic will increase nearly 1000x over the next decade. To meet that challenge requires cellular and Wi-Fi technologies to work together, according to New York University’s Ted Rappaport, the founding director of NYU WIRELESS, an academic research center combining engineering, computer science and medicine. During a panel discussion, he said Moore’s Law is about to catch up to wireless with huge bandwidths and 5G cellular around 2020. Rappaport is using NI’s software and products to work with millimeter waves and small future antennas that he said could be better than cellular technology.
Lost in Translation?
Perhaps the greatest challenge presented by the IoT, in which devices of all sorts talk to one another, is the communication itself. Before the timing and bandwidth hurdles come into play, the basic issue of one device understanding another must be tackled via standards and prototcols. (See “IoT: Talking the Talk” for some initiatives being pushed on the consumer side.) Because different communication technologies have different communication protocols and different industries have different standards, the communication challenge on the industrial side is daunting.
On an NIWeek panel, Intel’s general manager of its Internet of Things Group, Jim Robinson, compared the standardization effort to the early days of computing. “There was a ton of complexity to make the PC a simple platform,” he said, noting the various devices that had to connect into it and communicate with the computer. “It’s the same with the Internet of Things, except with super-fragmented markets.”
On the same panel, Sebastien Boria, Airbus’ mechatronic technology leader, said the company is already looking for more than just relaying information. “As we move forward, we need to ensure we don’t just have communication with the Internet of Things,” he said. “We also want a means to react to the information being communicated.”
That might sound like the cart leading the horse unless you’ve seen what Airbus is doing with its Factory of the Future initiative. Read more about it here.
Information needs to be collected and communicated before it can be reacted to. Likewise, devices need to be designed for there to even be an IoT. With all of the sensing and communications technologies that have to be integrated, simulated and tested on new products, the challenge of complexity comes up again and again.
It’s a safe bet that NI, which has come to be known for a platform that integrates hardware and graphical software, recognizes the advantage that could give engineers building the “things” that will combine hardware and software to connect the IoT. The new features of LabVIEW 2014 center around collecting and integrating real-time data, finding the data you need and visualizing it to reduce complexity. The visual aspect of LabVIEW is another key to reducing complexity, according to NI co-founder Jeff Kodosky.
“LabVIEW software tools provide the visualization that leads to better designs and higher productivity,” Kodosky said during his keynote presentation. “Visualization is a critical part of design generally, but visualizing cyber-physical systems (another term for the IoT) is a particular challenge. The physical part is straightforward but the cyber part is very difficult. Visualizing software is so challenging because of its abstract nature, unbounded flexibility and sheer complexity … Improving software design will have the biggest productivity impact.”
Despite the challenges ahead for cyber-physical systems, during the four days of keynotes, hundreds of technical presentations and expo floor demonstrations at NIWeek, it was obvious that NI and its partners believe the promise of the IoT is only a question of when, not if, it will come true.
For more information on NIWeek, visit National Instruments’ site and check out Desktop Engineering’s Twitter feed for our live coverage of the event. For more information on the Internet of Things, read the following features from Desktop Engineering’s August 2014 issue:
- Four Skills for the Internet of Things
- Feeling the Heat from Electrical Design
- Simulating Machine-to-Machine System Communication