Source: Raconteur
Internet of Things by the Numbers: Research Updates
IDC presented on August 4 its annual mid-year IoT review webcast, hosted by Vernon Turner, senior vice president and research fellow for IoT and Carrie MacGillivray, vice president of IoT & Mobile. Here are the highlights:
- An updated Digital Universe estimate of the amount of data created in the world annually (see above) forecast 180 Zettabytes (or 180 trillion gigabytes) in 2025, up from less than 10 Zettabytes in 2015 and 44 Zettabytes in 2020.
- Reaching the analytics phase of IoT: The actionable IoT Data–the IoT data that is analyzed and used to change business processes–in 2025 will be as big as all the data created in 2020. To make real time decisions, says IDC, “machine learning becomes important for the machine.”
- Growth rates: From 2020 to 2025, the volume of traditional data will grow by 2.3x; the volume of data that can be analyzed will grow by 4.8x; and the actionable data will grow by 9.6x.
- Connected devices: From less than 20 billion today to 30 billion in 2020 to 80 billion in 2025; by 2025, there will be 152,200 new connected devices every minute. “Everything we have of value will be connected to the internet,” says IDC.
- Current State-of-IoT in the U.S.: $230 billion will be invested in IoT in 2016 growing to $370 billion in 2018; 35% of U.S. companies are in the last 2 stages of IDC’s IoT maturity model; leading use cases are manufacturing operations, fleet management, and smart buildings.
- Market share of IoT networks in 2020: Wireless LAN/Wi-Fi 60%, Low-Power WAN 25%, Cellular 15%.
- The end of the self-built IoT platform? IDC thinks that there are between 300 and 400 company-specific IoT platforms. But they see a trend where companies abandon these efforts in favor of focusing on what they do really well. One of the winners this year has been Microsoft and its IoT platform with GE Predix coming to Azure.
- Other recent trends: Battle for Low-Power WAN market between proprietary solutions such as SIGFOX, LoRA and Ingenu and the ones favored by the cellular operators such as Narrow Band IoT; Applying AI to IoT security and a variety of IoT use cases; 2016 is the year of IoT developer—prominent example being IBM and AT&T’s announcement in July, coupling IBM Watson with AT&T development tools.
- Industry news: Cisco (IoT group moving organizationally from hardware to software) now has a well- known and established IoT platform, buying Jasper Wireless for $1.4 billion; Softbank buying ARM for $32 billion to take advantage of healthy growth of ARM’s market—their challenge is not to compromise ARM neutrality and fend off possible counter offers by Google or Microsoft.
IDC’s next IoT webinar, on September 22, will provide the results of its survey of 4,100 IoT decision makers in 25 countries.
In other research news, Machina Research released on August 3 its annual guidance on the size of the IoT market opportunity. Highlights:
- The total number of IoT connections will grow from 6 billion in 2015 to 27 billion in 2025.
- The total IoT revenue opportunity will be $3 trillion in 2025 (up from $750 billion in 2015). Of this figure, $1.3 trillion will be accounted for by revenue directly derived from end users in the form of devices, connectivity and application revenue. The remainder comes from upstream and downstream IoT-related sources such as application development, systems integration, hosting and data monetization.
- By 2025, IoT will generate over 2 Zettabytes of data, mostly generated by consumer electronics devices. However, it will account for less than 1% of cellular data traffic. Cellular traffic is particularly generated by digital billboards, in-vehicle connectivity and CCTV.
- China and the U.S. will be neck-and-neck for dominance of the global market by 2025. China which will account for 21% of global IoT connections, ahead of the U.S. with 20%, with similar proportions for cellular connections. However, the U.S. will still be ahead in terms of IoT revenue (22% vs 19%). Third largest market is Japan with 7% of all connections.
- Today 71% of all IoT connections are connected using a short range technology (e.g. WiFi, Zigbee, or in-building PLC), by 2025 that will have grown slightly to 72%. The big short-range applications, which cause it to be the dominant technology category, are Consumer Electronics, Building Security and Building Automation.
- Cellular connections will grow from 334 million at the end of 2015 to 2.2 billion by 2025, of which the majority will be LTE. 45% of those cellular connections will be in the ‘Connected Car’ sector, including both factory-fit embedded connections and aftermarket devices.
- 11% of connections in 2025 will use Low Power Wide Area (LPWA) connections such as Sigfox, LoRa and LTE-NB1.
Finally, ABI Research also sees machine learning playing an important role in the adoption of IoT by enterprises. It estimates that revenues generated by machine learning-based data analytics tools and services will reach nearly $20 billion in 2021 as Machine-Learning-as-a-Service (MLaaS) models take off.
Originally published on Forbes.com
The first mass deployment of driverless taxis will happen by 2020
Since the start of 2016, automakers, tech companies, and ride-hailing services have been racing to create a driverless taxi service. This service would mirror how an Uber works today, but there wouldn’t be a driver.
So far, the race has been brutal, as companies jockey for position by spending billions to acquire/invest in companies that will help make a driverless taxi service a reality. Uber recently took the pole position by announcing it would begin piloting its self-driving taxi service (with a driver still behind the wheel) in Pittsburgh later this month. But other companies, including almost every automaker, are quickly catching up as we reach the mid-way point in the driverless taxi race…
In a new report, we analyze the fast evolving driverless taxi model and examine the moves companies have made so far in creating a service. In particular, we distill the service into three main players: the automakers who produce the cars, the components suppliers who outfit them to become driverless, and the shared mobility services that provide the platform for consumers to order them.
Here are some of the key takeaways from the report:
- Fully autonomous taxis are already here, but to reach the point where companies can remove the driver will take a few years. Both Delphi and nuTonomy have been piloting fully autonomous taxi services in Singapore.
- Driverless taxi services would significantly benefit the companies creating them, but could have a massive ripple effect on the overall economy. They could cause lower traffic levels, less pollution, and safer roads. They could also put millions of people who rely on the taxi, as well as the automotive market, out of a job.
- We expect the first mass deployment of driverless taxis to happen by 2020. Some government officials have even more aggressive plans to deploy driverless taxis before that, but we believe they will be stymied by technology barriers, including mass infrastructure changes.
- But it will take 20-plus years for a driverless taxi service to make a significant dent in the way people travel. We believe the services will be launched in select pockets of the world, but will not reach a global level in the same time-frame that most technology proliferates.
How AI will change management
Interpersonal and judgment skills become more essential as AI evolves
AI will shift focus from coordination and control to judgment work, such as strategy and innovation, collaboration, people and community.
Network Map of Top 100 Cyber Security Influencers
We were very interested in seeing which Data Security influencers and brands were leading the online discussion, so we analysed tweets over the last 6 months mentioning the keywords: “data privacy” OR dataprivacy OR “data security” OR datasecurity OR datasec OR “data protection” OR dataprotection. We then identified the top 100 most influential brands and individuals leading the discussion on Twitter. What we discovered was a very engaged community, with much discussion between individuals and brands.
Startups Disrupting Healthcare with AI and Machine Learning
We identified over 90 companies that are applying machine learning algorithms and predictive analytics to reduce drug discovery times, provide virtual assistance to patients, and diagnose ailments by processing medical images, among other things.
A few investment highlights:
- Increasingly crowded imaging & diagnostics: 17 out of the 22 companies under imagining & diagnostics raised their first equity funding round since January 2015 (this includes 1st Seed or Series A rounds, as well as a first round raised by stealth startup Imagen Technologies). In 2014, Butterfly Networks raised a $100M Series C, backed by Aeris Capital and Stanford University. This was the third-largest equity round to AI in healthcare companies, after China-based iCarbonX’s $154M mega-round and two $100M+ raises by oncology-focused Flatiron Health.
- VCs invest in drug discovery: Startups are using machine learning algorithms to reduce drug discovery times, and VCs have backed 6 out of the 8 startups on the map. Andreessen Horowitz recently seed-funded twoXAR, developer of the DUMA drug discovery platform; Khosla Ventures and Data Collective backed Atomwise, which published its first findings of Ebola treatment drugs last year, and has also partnered with MERCK; Lightspeed Venture Partners invested in Numedii in 2013; Foundation Capital participated in 3 equity funding rounds to Numerate.
- Khosla Ventures backs 5 companies: Khosla Ventures has been the most active VC investor in the space, having backed 5 unique companies: California-based Ginger.io, which focuses on patients with depression and anxiety; healthcare analytics platform Lumiata; Israel’s Zebra Medical Vision and California-based Bay Labs, which apply AI to medical imaging; as well as drug discovery startup Atomwise.
- AI in oncology: IBM Watson Group-backed Pathway Genomics has recently started a research study for its new blood test kit, CancerIntercept Detect. The company will collect blood samples from high-risk individuals who have never been diagnosed with the disease to determine if early detection is possible. Other oncology-focused startups include Flatiron Health, Cyrcadia (wearable device), CureMetrix, SkinVision, Entopsis, and Smart Healthcare.
- Remote patient monitoring: New York-based AiCure raised $12.3M in Series A funding from investors including Biomatics Capital Partners, New Leaf Venture Partners, Pritzker Group Venture Capital, and Tribeca Venture Partners, for the use of artificial intelligence to ensure patients are taking their medications. California-based Sense.ly has developed a virtual nursing assistant, Molly, to follow up with patients post-discharge. The company claims Molly gives clinicians “20% of their day back.” Sentrian, backed by investors including frost Data Capital, analyzes biosensor data and sends patient-specific alerts to clinicians.
- Core AI companies bring their algorithms to healthcare: Core AI startup Ayasdi, which has developed a machine intelligence platform based on topological data analysis, is bringing its solutions to healthcare providers for applications including patient risk scoring and readmission reduction. Other core AI startups looking at healthcare include H2O.ai and Digital Reasoning Systems.
The U.S. Artificial Intelligence Market
The New Artificial Intelligence Market by Aman Naimat, published by O’Reilly:
There are only 1,500 companies in North America that are doing anything related to AI today, even using its narrow, task-based definition. That means less than one percent of all medium-to-large companies across all industries are adopting AI.
Big Auto Self-Disruption
Traditional automotive OEMs have begun making deals at a frantic place, seeking to remedy their shortcomings in auto tech and ride-hailing disciplines. Using CB Insights data, we mapped out the key auto tech partnerships, investments, and acquisitions of these corporations over the past three years.
We focused on auto OEMs’ private markets activity within our definition of auto tech, which includes startups that using software to improve safety, convenience, and efficiency in cars (and excludes activity in fields such as energy/powertrain, parking, and rentals/marketplaces). We also looked at their major engagements with ride-hailing companies and large tech corporations.
Scanning the timeline, the acceleration of activity seen in 2016 is immediately obvious.
[youtube https://www.youtube.com/watch?v=CMYlKBDsDtc]
Google 1 Yahoo 0
Many of the obituaries for Yahoo have contrasted its demise with the flourishing of Google, another Web pioneer. Why was Google’s attempt to “organize all the world’s information” vastly more successful than Yahoo’s? The short answer: Because Google did not organize the world’s information. Google got the true spirit of the Web, as it was invented by Tim Berners-Lee.
In his book Weaving the Web, Tim Berners-Lee writes:
I was excited about escaping from the straightjacket of hierarchical documentation systems…. By being able to reference everything with equal ease, the web could also represent associations between things that might seem unrelated but for some reason did actually share a relationship. This is something the brain can do easily, spontaneously. … The research community has used links between paper documents for ages: Tables of content, indexes, bibliographies and reference sections… On the Web… scientists could escape from the sequential organization of each paper and bibliography, to pick and choose a path of references that served their own interest.
With this one imaginative leap, Berners-Lee moved beyond a major stumbling block for all previous information retrieval systems: The pre-defined classification system at their core. This insight was so counter-intuitive that even during the early years of the Web, attempts were made to do just that: To classify (and organize in pre-defined taxonomies) all the information on the Web.
Google’s founders were the first to seize on Berners-Lee’s insight and build their information retrieval business on tracking closely cross-references (i.e., links between pages) as they were happening and correlate relevance with quantity of cross-references (i.e., popularity of pages as judged by how many other pages linked to them). This was what set Google apart from its competitors, including Yahoo. Having a so-called “first-mover advantage” (yet another example that there are no universal “business laws”), Yahoo worked hard and employed many people in organizing in a neat taxonomy the rapidly-growing content of the Web. It even had a Chief Ontologist on staff.
Google’s ranking system gave you the best of both worlds. Yahoo was a card-catalog of the web, letting you effectively search for the right “books” based on what they were titled. Google’s system let you search through all the pages of all the books in the entire library. It was far more comprehensive, plus it still managed to get good stuff to the top of the list.
Berners-Lee’s insight is frequently linked to Vannevar Bush who wrote in 1945, “Our ineptitude at getting at the record is largely caused by the artificiality of systems of indexing… Selection [i.e., information retrieval] by association, rather than by indexing may yet be mechanized.” But I prefer to start the history of the Web (and organizing information) with what was, to my knowledge, the earliest use of cross-references.
This was Ephraim Chambers’ Cyclopaedia, published in London in 1728. While lacking the worldwide platform for “crowd-sourcing” references that Berners-Lee invented, Chambers shared with him (and Bush) a dislike for hierarchical, alphabetical, indexing systems. Here’s how Chambers explained in the Preface his innovative system of cross-references:
Former lexicographers have not attempted anything like Structure in their Works; nor seem to have been aware that a dictionary was in some measure capable of the Advantages of a continued Discourse. Accordingly, we see nothing like a Whole in what they have done…. This we endeavoured to attain, by considering the several Matters [i.e., topics] not only absolutely and independently, as to what they are in themselves; but also relatively, or as they respect each other. They are both treated as so many Wholes, and so many Parts of some greater Whole; their Connexion with which is pointed out by a Reference. So that by a Course of References, from Generals to Particulars; from Premises to Conclusions; from a Cause to Effect; and vice versa, i.e., in one word, from more to less complex, and from less to more: A Communication is opened between the several parts of the Work; and the several Articles are in some measure replaced in their natural Order of Science, out of which the Technical or Alphabetical one had remov’d them.
Chambers’ Cyclopaedia was the earliest attempt to link by association all the articles in an Encyclopedia or, in more general terms, of everything we know at a given point in time. And like the World Wide Web, it moved some people to voice their concern about what Google is doing to our brains. The supplement to the 1758 edition of the Cyclopaedia says:
Some few however condemn the use of all such dictionaries, on the first pretence, that, by lessening the difficulties of attaining knowledge, they abate our diligence in the pursuit of it; and by dazzling our eyes with superficial shew, seduce us from digging solid riches in the mine itself.
The fear of what tools for organizing information could do to our thinking (and livelihood) was renewed many-fold with the advent of modern computers. “They can’t build a machine to do our job; there are too many cross-references in this place,” says the head librarian (Katharine Hepburn) to her anxious colleagues in the research department when a “methods engineer” (Spencer Tracy) is hired to “improve workman-hour relationship” in a large corporation. By the end of the film, Desk Set (released in 1957), she proves her point by winning, not only the engineer’s heart, but also a contest with the ominous looking “Electronic Brain” (aka Computer).
Automation—replacing librarians and their card catalogues—has been at the heart of Google’s success and obsession with “scale” (and “at scale” has become an obsession for Silicon Valley). But this automation has led to augmentation, to supporting our thinking by creating a new way to organize the world’s information, one that is more in line with our thought process and more in line with the impossible-to-catalogue current volume of (valuable and useless) information. As Vannevar Bush wrote:
The human mind… operates by association. With one item in its grasp, it snaps instantly to the next that is suggested by the association of thoughts, in accordance with some intricate web of trails carried by the cells of the brain … One cannot hope to equal the speed and flexibility with which the mind follows an associative trail, but it should be possible to beat the mind decisively in regard to the permanence and clarity of the items resurrected from storage.
Originally published on Forbes.com