Happy 2017. We’re at the beginning of another exciting year for IoT. Last year, we’ve seen continued interest & ever expanding applications or solutions for IoT. Before we look at 2017, let’s look back at last year’s trends & which were spot on & which ones were off-the-course.
- Big Data becomes mainstream (with IoT) BDA overall had a mixed year. Many BDA projects took longer to realize their full potential due to the lack of amount of data to create a perfect prediction model or ,platforms that are designed to handle large volumes of data that are under utilized. In a nut shell, BDA is used on trying to solve problems that has “Little” data to begin with. The other other factor is the amount of work in curating & formating the data before it can be used (which are still done manually). A case to point is the failure of BDA tools used in polls to predict the possible outcome of the US Elections which is an event that should have generated massive amount of data. Is Donald Trump’s Surprise Win a Failure of Big Data?
- SOM adoption into more devices It was quite an exciting year with M&A’s in the semiconductor industry. The company, whose chips were used in the arduino platfrom, Atmel was acquired by Microchip. Freescale acquired by NXP. And shortly, it was announced that NXP itself is in the process of being acquired by Qualcomm. Avago merges with Broadcom. Avago also acquired LSI. Japanese internet and telecommunications conglomerate SoftBank Group Corp, bought ARM Holdings PLC, the U.K.-based designer of the microprocessors that power more than 95% of the world’s smartphones. The IoT hardware embedded designer will have a tough time in the selection of silicon suppliers for their product designs. With the mergers & acquisitions, we have very large companies that could deliver the components for end-to-end IoT hardware solutions, even smaller SOM (10 nm process?) & longer battery life devices. Most of the M&A’s are seen as pretty strategic approaches to gain IoT market-share.
- Social IoT network In 2016, the founder of the most popular social network, Mark Zuckerberg announced on his side project to implement IoT as a home automation proof-of-concept the future of IoT where FB will be part of that vision. In 2016, the focus was on bots(automated software that are used to provide interactive messaging & feedback to humans) and voice activated assistants devices with AI such as Alexa (Amazon) & Google Home, instead of interacting with your smartphones with messaging/texting, interactions will even be easier with voice. Probably in the not too distant future , you can find yourself talking to everyday appliances or “Things” like a desk or a mirror (remember the snow white fairy tale, where the evil queen converse with the magic mirror?)
- 5G still have not seen an actual commercial roll out yet. However, efforts to reduce the latency of response time (major potential of what 5G can bring to IoT applications) are being mitigated by companies who are running analytics closer to the devices using fog computing as opposed to cloud computing or using hardware acceleration like GPU’s or FPGA’s at the server side. Nevertheless, 5G will eventually deliver the bandwidth & “real-time” response to enable powerful IoT applications.
- IoT Architect Developer With the need to develop IoT products & services at an accelerating rate, the demand for IoT specialists are as high or even higher in the future. With the recently large DDS attacks through IoT devices, hackers are starting to target IoT devices as well. The IoT developer will need to add cyber-security skills on top of multi-domain knowledge in software & embedded hardware to design a reliable IoT solution.
What’s in store for 2017?
AI becomes mainstream (with IoT) Last year, starting with the announcement of Google’s Deepmind beating one of the world’s best human GO player & new processor designed for A.I. and further announcements on ‘Partnership on AI’ formed by Google, Facebook, Amazon, IBM and Microsoft , A.I. (Artificial Intelligence) seems to “jump” straight from science fiction movies & our imagination to mainstream usage almost immediately. Combining with voice & other forms of interactive inputs (brain wave sensors?), will we see IoT devices that can be used by non-tech savvy consumers.
5G Shipping consumer products using 5G technology will appear by end 2017. With the sheer number of “Things” to be connected, higher bandwidth will be number one priority along with very fast latencies. With the bottleneck removed, we do not need to have a lot of processing power at the edge, giving devices weeks or years of operations on batteries. Don’t be surprise if the first wave of 5G-powered products are not smartphones but rather simpler IoT devices.
IoT Platforms We have seen many many IoT platforms emerging & making it easier for companies to develop IoT solutions quickly. IoT platforms are emerging as the central backbone of these IoT deployments – By 2019, the IoT Platform market will reach $1 billion. Most platforms have they own proprietary API , libraries & security etc. where development done for one platform does not work at at if migrating to another. Also, there’s the issue of no-one-size-fits-all IoT application. Will we see an innovative IoT platform that allows developers to create solutions that allows inter-operability or tight integration of different components of multiple platforms? An example of IoT platform
SOC With the recent M&A in the semicon industry, we will see chip designs where more functionality is integrated like including sensors(MEMs), micro-controller & wireless front-end in the same chip. This can possibly enable smaller products, longer batter-life & lower costs.
Cloud There would be a renewed focus on cloud computing/fog computing, FPGA/GPU accelerated microservices or Coud-within-a-Cloud Docker services , security & ID technologies used in IoT. With other resources intensive application areas such as Augmented Reality & blockchain, AI etc. we will need a major shift on utilizing cloud for IoT solutions
In-door geo-location In the IoT context, anything connected to the internet will automatically publish it’s geo-physical location as part of asset tracking or management. As such, humans are the already automatically tracked already through their smartphones. However, most geo-location technologies are based on GPS & needed open skies to work properly & does not work well or at all inside buildings. Therefore, other technologies are needed to work indoors. Expect sub-cm accuracy technologies appearing in the horizon that enable revolutionary applications in the the industry.
2017 will shape up to be a very interesting year for IoT on all fronts combining technologies in A.I. & A.R. & cyber-physical systems.