What lies ahead will likely continue to transform our lives in unimaginable ways. While some predictions may take a bit longer to materialize than 2017, they provide a solid sense of emerging technology and engineering trends in the New Year and beyond.
So tell us—what are your predictions for the coming year?
By Cees Links, General Manager of the Wireless Connectivity Business Unit, Qorvo
We think that 2017 will be a time of evolution for the smart home and the IoT. Over the last few years, most device developers and manufacturers were learning how to successfully web-connect their devices—how to make their devices and appliances accessible to smartphones and tablets over a web-interface. The next step of the evolution is about to arrive, including making these devices work with each other, giving them intelligence, and making the entire system easy to use.
Security applications should be able to talk to lighting, energy management, and health applications. Data needs to be shared, not just between the sensor devices themselves, but also between the various controllers and actuators. In addition, sensing edge nodes need to talk to the various applications. For example, a light motion detector should be able to provide data on movement in the home to the security system, ensuring that all “improper” motion is detected and reported. The same motion detector can serve to turn lights on and off as a person walks through the home. In addition, it can sense when someone enters and leaves a room and adjusts the climate controls accordingly, keeping the home comfortable while at the same time conserving energy.
Finally, smart home systems need to actually become smart. Much like the Amazon Alexa and Google Home, which transfer data to cloud intelligence for analysis and response, connected home devices need to connect to the cloud to learn how the household functions and to evolve services to better meet the needs of the residents—ensuring their comfort, security, and optimal efficiency. In many ways, the smart home will mature into a smart home butler, anticipating needs and making lives more comfortable and efficient.
The smart home is evolving into a consolidated service with smart appliances and devices throughout the home. However, in contrast to today’s siloed connected devices (each with its own method of control, online dashboard or smartphone app), the industry will come together and offer unified smart services with a single, easy-to-use dashboard with a common look and feel across all services—be it security, health, e-commerce, climate control or energy management. All will be obvious and easy to use.
Ease of use, intelligence, and an ability to finally converse and work together is both the challenge and opportunity for the smart home that 2017 is offering the IoT development community.
By Patrick McFadin, Chief Evangelist for Apache Cassandra, DataStax
The Emergence of the Data Engineer
The term “Data Scientist” will become less relevant and be replaced by “Data Engineers.” This may seem like a slight change, but it is an important one. Historically, a Data Scientist analyzed data in a reactive manner, such as analyzing company data and making recommendations based on missing data and/or data trends. However, as more data, especially from IoT devices, is being embedded into applications, Data Engineers will proactively use the data to improve existing technologies and services, leading to potential innovation across business units and industries. The current trend in stream analytics means that engineers will be looking at near-real time data for immediate action.
Security: Blurred Lines in IoT
The IoT’s growth has largely gone unchecked. With a lack of standards and an explosion of data, it isn’t entirely clear who is responsible for securing what. Most at risk are ISPs, which is why these providers will take a leading role in the security conversation in the year ahead and even begin blocking traffic that doesn’t meet requirements. Additionally, device makers, who have historically left security to others in the ecosystem, will begin feeling the pressure to build additional security directly into devices.
The Software License: A Victim of the Cloud
Much has been said about the death of the traditional software license, and you can point your finger in blame at the emergence of the cloud. It is true that traditional software companies need to rethink how they sell, but there are unforeseen impacts as well. For instance, the Big Data open source community must also contend with the decline of licenses, making it difficult for those looking to show a clear path to profitability. Cloud is, as they say, the great equalizer—much to the dismay of those trying to make money.
By Colin Geis, Director of Product Management – IIoT, Red Lion Controls
In 2017, the Industrial Internet of Things (IIoT) will move from a “nice to have” to mainstream initiative. Many more companies will make the move from talking and thinking about the IIoT to actually implementing it. However, one IIoT challenge they will face is properly deploying the networking equipment required to get data to the cloud. We will also see a shortage of tech-savvy staff who are knowledgeable and trained on IIoT technology. This will make easy-to-use products with remote monitoring capabilities important for companies that lack technical staff.
By Liam Devlin, CEO, Plextek RFI
2017 will see continued growth in the development and evaluation of mm-wave 5G communications, as well as a growing interest in drone detection and jamming. The cost and performance of commercially available drones represent a potential security threat to airports, power stations, and other critical infrastructure. High-resolution radar systems are the key to detecting small, highly maneuverable drones, and GaN technology can provide adequate power to jam the aircraft’s control links. High-performance, low-cost GaN transistors, capable of generating adequate power across the relevant bands, are now readily available.
Many manufacturers are investing heavily in developing solutions for the Internet of Things (IoT). The interest stems from the potentially huge volumes in which IoT products will ultimately be deployed, but in fact, the IoT is already with us. IoT solutions are currently being deployed and actively used today in wireless street lights, fleet management, vehicle tracking, healthcare, smart homes, and a host of other M2M and Machine Type Communications (MTC) applications. These applications currently use a range of personal-area connectivity technologies including Bluetooth, ZigBee, proprietary ISM band solutions, and local-area networks like WiFi 802.11ah (HaLow). They will also increasingly start to use Low Power Wide Area (LPWA) technologies including LoRa, SigFox, Telensa, and LTE-based standards such as NB-IoT.
Low-cost, high-volume IoT components and products based on these emerging IoT standards will see considerable development activity in 2017 and will begin to be launched into the market. However, there will be no single solution for IoT applications; many of the existing solutions will continue to grow alongside the newer technology standards that are joining the global M2M connectivity behemoth.
By R. Dale Lillard, President, Lansdale Semiconductor, Inc.
I believe the 2017 economy will be very similar to 2016, which is defined as “slow is normal.” The Semiconductor Industry Association has forecasted little growth for 2017, and I believe their prediction is correct. Many of our customers continue to be conservative with inventory builds. They only buy what they absolutely know they have orders for, even if it costs more in price and cycle time. Beyond 2017, the new administration will be good for business due to Trump’s suggested tax reductions for defense contractors and increased military spending. It will take at least a couple of years for that improvement to trickle down to us, but I expect sales to improve towards the end of 2018 and in 2019 as the changes start to impact the economy.
By Mark Murphy, Senior Director, Marketing and Business Development, RF Power, MACOM
The accelerating mainstream adoption of high-efficiency, low-cost Gallium Nitride (GaN) for commercial markets heralds a major disruption across a host of applications today that rely on legacy magnetron technology as a heating and lighting source. Vacuum tube-based magnetrons will increasingly be displaced by GaN-based, solid-state solutions that leverage RF energy to control electromagnetic radiation with semiconductor-caliber reliability and enhanced precision, while simultaneously enabling smaller form factors.
One key, near-term target application for RF energy is the traditional microwave oven. With an RF transistor generating precision-controlled energy fields, microwave cooking will be radically transformed. Microwave ovens will be programmable to target specific areas with varying amounts of energy, ultimately enabling more thorough, efficient cooking. This technology shift is already underway in industrial cooking applications and will penetrate the mainstream consumer market thereafter.
Plasma lighting is another key target application for RF energy. Plasma lights provide key advantages over alternative mainstream light sources. They’re much smaller in size than the conventional fluorescent and high-intensity discharge (HID) lamps commonly used today to light large areas, and they provide much higher luminaire efficiencies. These qualities enable plasma lights to deliver high levels of visual acuity while requiring fewer light fixtures and consuming less power.
Looking beyond 2017, RF energy holds the promise to transform applications spanning from automotive ignition to industrial, scientific, and medical (ISM) applications including material drying, blood and tissue heating and ablation, and beyond. The RF devices that underpin these systems must strike an optimal balance of performance, power efficiency, small size, and reliability at a sub-LDMOS price point suitable for mainstream commercial adoption. Fourth generation GaN on Si (Gen4 GaN) is ideally suited for this purpose.
By Raj Radjassamy, Senior Product Manager, GE Energy Connections
Looking ahead to 2017 and beyond, we predict major trends from both end-application and advanced technology perspectives. At the end-application level, we see continued growth of the industrial internet, home automation, and deep learning—all enabled by advances in big data and expanding data center capacity. All these technologies support the ever-expanding demands of data mining and related data collection, management, and analysis.
2017 also marks a point of momentum for a number of core technologies important to our customers including: high-voltage direct current (380-volts DC) architecture, liquid cooled power supplies for data centers and industrial applications, and lower supply voltages for field programmable gate array (FPGA), system-on-a-chip (SoC), and microcontroller unit (MCU) technologies.
We also see a confluence of several related technology trends—networking, computing, storage, security, Infrastructure, and mobility—impacting software-defined everything (SDx) that will reduce total cost of ownership and maximize the return on investment for our customers.
By Kin-Yip Liu, Senior Director of Solutions Architecture and Segment Marketing, Cavium
As the industry prepares for initial 5G deployments and 5G-enhanced viewer experience at the 2018 Winter Olympics in South Korea, we expect that in 2017 there will be significant industry focus and momentum in developing pre-5G and 5G solutions, accelerating network functions virtualization (NFV) performance and efficiency, and driving much enhanced security.
Service providers have been in a long journey since 2012 to adopt NFV to improve operational efficiency and enable agile services deployment. So far, NFV deployments are still limited. For many network functions, significant gaps exist in performance and cost efficiency, when comparing NFV implementations using general purpose hardware against purpose built appliances. NFV implementations may require many more general purpose hardware resources to deliver acceptable performance. This drives up cost, especially in power consumption. Hardware acceleration with standard software API and MANO (management and orchestration) interface is a plausible solution. ETSI NFV ISG has included acceleration content in the NFV 2.0 specifications. In addition, a new ETSI NFV proof of concept “network function acceleration with resource orchestration” illustrates multiple use cases where network functions are accelerated using hardware accelerators, which are managed by the NFV MANO system.
C-RAN (Cloud or Centralized Radio Access Network) is one of the most complex network functions in terms of NFV adoption. This is because general purpose processors are not effective for processing major portions of wireless standards, especially the wireless physical layer and the unique air interface cryptographic algorithms. As such, there will be a gradual migration where the near-term focus is to boost performance of base stations to deliver 5G throughput and, eventually, efficient C-RAN solutions. With C-RAN implementations, the baseband processing is virtualized and centralized at COTS servers in the telco edge/cloud.
In addition to boosting throughput, C-RAN requires multiple other innovations. To maximize 5G deployment opportunities, the front haul, which is the switching fabric between remote radio units at cell sites and the centralized virtual baseband, needs to support multiple split options of the LTE stack processing with standardized API. Mobile and multi-access edge computing (MEC) promises much improved latencies and mobile user experience, as well as exciting new services and business opportunities. Network slicing enables dynamic orchestration of logical network slices, which are individually optimized for various 5G new services.
As IoT and 5G enable connected devices on a massive scale, security requirements and complexity increase tremendously. Effective end-to-end security solutions are a major pre-requisite for enabling effective IoT solutions.
By Jun Ye, Co-Founder and President of Pilot Labs Moorebot
Artificial Intelligence (AI) is coming. When its cost drops down to levels that can produce actual business benefits, it will explode into many areas. More interactive devices will serve as Natural User Interfaces (NUI) for smart homes and other automated applications. In addition to providing big improvements in voice interaction, NUI technology, which is based on deep learning, also combines facial, gesture, and expression recognition and interaction history. It will become the control interface that “knows you best.”
By Brent Ward, VP Business Development, Device Solutions Inc. (Cellio)
For the Industrial Internet of Things (IIoT) we see the following top three trends dominating the landscape and driving increase adoption of technology “in the field”:
- Leveraging LTE: We see LTE (CAT1) driving the costs and efficiencies of devices deployed to the field dominating the conversations, research, and planning for large enterprise deployments.
- Unified Dashboards: We see the demand expressed in 2016 driving organizations and vendors to provide either devices that can provide data that be shown on any dashboarding service or vendors offering dashboards that can integrate data across the wide variety of devices that are currently deployed, being deployed, or will be deployed in the future such that they have a unified view of all of their devices.
- Integrated Ecosystem Offerings: We have seen an increasing demand by customers to get into IIoT with “drop in” solutions that guarantee their success in deployment for their IIoT implementations but also provide flexibility to “plug and play” with any of the ecosystem elements as their organizations develop the infrastructure needed to fully leverage all the information and data collected throughout their deployed devices.
- Bonus: Backlash on Too Much Data: We have heard an increasing angst coming from clients who have seen vendors try to drown them with too much data. We predict that customers will increasingly demand less data, more targeted, and expect that it comes to them on an exception basis. They do not want to create another job just to interpret the data that is supposed to helping them automate their improvements in efficiency and quality.
Overall, our prediction is that 2017 will see the greatest adoption of IIoT technologies yet. The convergence of lower costs devices, lower cost sensors, lower cost wireless services, lower cost data storage/dashboarding/analytics, and ultimately lower solution cost is leading to an unprecedented set of conditions to make ROI’s and Break-even’s possible where they weren’t before. Interoperability and implementation of agreed upon standards will accelerate the lessening of concerns about risks putting things in place to gain the competitive advantages inherent in IIoT.
By Debbie Greenstreet, Marketing Manager Processors, Texas Instruments
Gone are the days of using our fingers to give our devices a command. 2017 will be the year that household appliances, electronics, and many other devices in our environment are all centrally controlled by speech. Voice will be the user interface.
While we have seen this paradigm shift start to take place with applications from Amazon and Google, we have yet to see devices truly be voice activated. Meaning your oven can turn on with the simple phrase “oven turn to 350 degrees” or your thermostat can change temperatures with a simple: “thermostat, change the temperature to 70 degrees.”
Continued rapid advances in real time speech recognition algorithms, coupled with cloud based speech processing will enable a wide range of voice-based products. Similarly, efficient packaging of audio pre-processing functions (those processes that help separate voice from the rest of the audio in the area) for “far field” applications will open up further opportunities for innovation. As such, we will most likely see the following:
- Speaking directly to our appliances. Yes, while there are already gateway based solutions today that are addressing functions like turning on lights and music, we will start to see speech recognition features appear in stoves, refrigerators, thermostats, etc. such that we can control functions verbally like set the temperature to 70 degrees. And, this capability won’t be limited to purely consumer use. Expect to see more voice activated functionality in building automation applications like alarms, entry locks, as well.
- Voice commands in our automobiles. We have gotten used to our cars speaking to us, it shouldn’t be a surprise that we will in turn be able to control functions in our vehicles, hands free. Consider the safety improvement of commanding on the flashers or high beams in a challenging situation versus struggling in the dark to find the appropriate knob.
- A surprise. Well I don’t suggest knowing what the answer is to that yet but I predict that there will be some exciting, creative application to which we will say, of course, finally, why not? And there is a good chance that there will be more than one.
Bill Balmer, Manager, Global Business Development, Ethernet and NFV markets, ADVA Optical Networking
Encryption will be hugely significant in 2017 as the cloud expands its dominance in enterprise networks. IPsec’s excessive use of bandwidth and added latency will create an environment where encryption begins to impact services. As enterprises migrate toward the cloud, the bandwidth price of 40% for legacy encryption techniques will dramatically increase service costs. What’s more, added delays for slow encrypting times can more than double the latency of distributed compute applications.
The explosion of the number of internet of things applications, each using individual IPsec encryption, will compound the problem with many instances of unique encryption sessions operating on cloud links. Most of these apps will not be tested for security strength and will not be coordinated into any corporate managed key system. Each will add extra burden to network bandwidth requirements.
SD-WAN deployments will create more requirements for on-the fly protection with more network traffic travelling over multipath topologies.
New legal threats for data protection like the European Union General Data Protection Regulation now extend the reach of national liability laws across the globe. Organizations that treat protecting customer data as a best-effort practice will be doing so at their own risk.
All of these factors will create the need for a prophylactic approach to create secure cloud access.
Hardware-based encryption techniques that have been used in specialized networks are starting to be deployed in enterprise and EaaS services. This trend will increase in 2017 as these devices offer network-grade encryption with negligible latency and bandwidth requirements drastically below the pricing of traditional encryption appliances.
Virtualization of encryption will also be a growing trend as enterprises look for cost-effective ways to leverage high-quality encryption. Encryption techniques that used to be only affordable to governments and Fortune 500 companies will now be available through subscription services.
By Chuck Delph, President, Avnet Electronics Marketing Americas
When I look at the year ahead, I think the Internet of Things (IoT) will continue to dominate our conversations, strategies, product roadmaps, business plans – and your headlines. It’s a significant driver for the electronics industry. In fact, I’d say we’ve only scratched the surface of defining “50 billion connected devices,” what that really means and how we fit in that universe. Companies are struggling to define their path to success. The opportunity is so huge, how do you capitalize on it? Is it hardware or software? The cloud? Will the real surprise winners end up coming out of nowhere?
It’s because of that uncertainty that I expect 2017 to be the year we’ll see more interesting IoT alliances come together. I see more companies building partnerships as “traditional” players join forces with more nontraditional companies and explore multiple models. You’ll see more deals like Avnet’s acquisition of Hackster.io, as companies experiment with different formulas. We’re all looking to define success in IoT; and there are so many different paths to take. While I don’t know which formula will win in IoT, I do know one thing for sure: as an industry, we’re not slowing down.
Ethernet Alliance - Board of Directors
In 2016, Ethernet’s ongoing expansion continued to broaden at an unprecedented rate, impacting sectors from hyperscale data centers, enterprise, cloud computing, automotive, Internet of Things (IoT), and Power over Ethernet (PoE), among others. As the leading industry voice for Ethernet, the Ethernet Alliance is on the front lines of the technology innovations that will carry Ethernet through 2017 and beyond. Here, three of the Ethernet Alliance’s key experts offer their views on what the New Year might hold:
By George Zimmerman, Technical Chair, Ethernet Alliance; President, CME Consulting
In 2017, the maturation of PoE will accelerate the development of uses beyond IP phones and wireless access points, to include lighting, sensors, and Internet-of-Things applications. Wired Ethernet will continue to grow as an enterprise infrastructure application, and that includes combining data with power. Power over Ethernet use, already over 15 years old, will accelerate, providing DC power over 4 twisted pairs of copper at increased levels and more efficiently than before, in anticipation of the maturing IEEE 802.3bt PoE standard. With the increasing variety of PoE types and devices, efforts of groups such as the Ethernet Alliance will help guide consumers in selecting compatible products.
By Greg McSorley, Ethernet Alliance Board of Directors; Technical Business Development Manager, Amphenol
On the cabling front, the use of passive copper cables in the Data Centers employing top-of-rack applications will grow as opposed to shrink in the coming years. This is due in large part to the cost of 25Gb, 50Gb and 100Gb transceivers and active components needed for the Active cables as well. In applications where the link lengths are less than three meters, passive copper cabling is much more cost effective and power efficient. Compared to even Direct Attach Cable (DAC) and Active Optical Cable’s (AOC), passive twin-ax cabling is much more attractive given that it doesn’t have any of the electronics that you would need for the standard DAC. This lowers the cost of cable alone by 30%. The AOCs are about five to six times more expensive. While the use of two transceivers and a fiber jumper in between is 10 times more expensive.
By David J. Rodgers, Ethernet Alliance Board of Directors; Sr Product Marketing Manager, SAN Protocol Tools, Teledyne LeCroy
The emergence of 25/50/100Gb high-speed Ethernet rates will change the test and measurement landscape for the development and support communities. Hardware engineering teams developing product for the newer Ethernet speeds (ex. IEEE 802.3bj and IEEE 802.3bs) will need to become “protocol aware” as improvements in PHY layer dialog for auto-negotiation and the advent of transmitter training are supported. The codependent interoperation of physical connections and upper layer transactions will be even more intertwined for the successful application and deployment of solutions offering new data rates. Early adopting end-users too will see a need to revamp the tool chest used for supporting their communities. Relationships between hardware vendors and their chosen Test and Measurement tools and vendors will percolate through the development cycles and be leveraged into the support channels to ensure rapid and complete problem remediation. Improved channels of communication and more concise descriptions of issues will occur between the IT/DC manager and their product vendors. All participants in the Ethernet ecosystem will gain greater knowledge about and lean more heavily on the hardware and software Test Tools community to safeguard the investments in equipment and the quality of service each supply.