Roadmap for the Trillion Sensor Universe

Roadmap for the Trillion Sensor Universe
Dr. Janusz Bryzek
VP Development, MEMS and Sensing Solutions
Fairchild Semiconductor
Hayward, CA
iNEMI Spring Member Meeting and Webinar
Berkeley, CA, April 2, 2013
Mobile Sensor Tornado
•
•
Introduction of iPhone and WII in 2007 started the mobile sensor Tornado.
Between 2007 and 2012, mobile sensor consumption grew from 10M units to
3.5 billion units.
Four sensors dominate the volume:
•
–
–
–
–
•
Microphone
Acceleration
Magnetic
Gyro
Large number of additional sensors has been embedded in cell phones and
more is coming.
Large number of sensor based accessories emerged for mobile devices
creating a new category “Appcessories”.
•
– App with an Accessory.
2
Sensor based Appcessories, Sampler 1
Sleep monitor
Vital signs monitor
3
Sensor based Appcessories, Sampler 2
Smart Shoe
Nike Fuel Band
Smart pedometers
http://www.kickstarter.com/projects/597507018/pebble‐e‐
paper‐watch‐for‐iphone‐and‐android
Baolab tennis racket
4
94Fifty sensor basketball
Smart surfboard with GPS and IMU
Sensor based Appcessories Sampler 3
Augmented Reality
(Google Glasses)
Headset with brain interface
Recon Goggles with heads-up display, GPS and IMU for skiers
5
Sensor based Appcessories, Sampler 4
•
Lapka: gathers:
–
–
–
–
–
•
Humidity
Temperature
Radiation
Electromagnetic frequencies (EMF)
Organicity (checks for nitrates, which are commonly
used in chemical fertilizers).
Node: includes
–
–
–
–
–
–
–
–
•
Motion sensors for wakeup
Temperature gauge
Barometer
Ambient light
Humidity
Point-and-shoot temperature sensor
Color sensor for matching colors
Forthcoming range of gas sensors.
http://www.fastcodesign.com/1670479/iphone‐
sensors‐test‐if‐your‐food‐really‐is‐organic#1
http://www.tuaw.com/2012/12/11/node‐gives‐
your‐iphone‐sensory‐input/
6
Where the Visible Growth is Coming from?
Yole 2012
7
Emerging Mobile MEMS Products
Yole 2012
8
No Trillion Sensors Mentioned by Market Research Companies…
Although there are
Visible Pointers to Trillion Sensors
Abundance/XPrize
•
Concept of Abundance was introduced by Peter Diamandis of XPrize Foundation in his book “Abundance”. •
•
•
Peter believes that Abundance is feasible in just one generation (20 years) due to emergence of exponential technologies delivering a faster growth of supplies and services than the global demand. Eight exponential technologies are considered critical to achieve abundance:
•
•
•
•
•
•
•
•
•
Defined as equality between supply and demand for goods and services for all people on Earth. Biotechnology and bioinformatics
Computational systems
Networks and sensors
Artificial intelligence
Robotics
Digital manufacturing (3D printing) and infinite computing
Medicine
Nanomaterials and nanotechnology.
Abundance envisions the need for 45 trillion networked sensors in 20 years.
http://www.abundancethebook.com/
10
Smart Systems
•
Harbor Research introduced a concept of Smart Systems in the Era of Pervasive Internet.
•
Defined as a fusion of computing, communication and sensing.
 People, devices, sensors and businesses are connected and able to interact. •
Harbor Research leading markets for smart sensing systems include:
•
•
•
•
•
•
•
Cell phones
Health monitoring devices
Smart grid infrastructure
Automotive
IT
Industrial systems
Smart business will enable collective awareness, creativity and better decision making capabilities, driving the largest growth opportunity in the history of business.
http://www.harborresearch.com/Home.htm
11
Qualcomm Swarm Lab at UC Berkeley •
•
Swarm Lab projects 1000 radios per person on Earth by 2025, with trillions of connected devices as the swarm around the edge of the Cloud, to link the cyber and physical/biological worlds.
Trillions wireless devices creates a challenging demand for RF bandwidth.
–
–
Swarm Lab is developing next gen low power radios enabling bandwidth sharing.
IPv6 introduced in June 2012 paved a way for sensory swarms.
•
•
•
12
Increased IP addresses from 232 = 4.3 billion to 2128, or 340 undecillion, or…
340 trillion trillion trillion, or…
340,282,366,920,938,463,463,374,607,431,768,211,456 Bosch Sensory Swarms
•
Bosch presented a vision for 7 trillion devices consisting of Sensory Swarms connected to the Internet to serve 7 billion people by 2017.
•
•
Funded development group.
This vision translates to 1000 sensors per average person. •
Current applications supported by large number of sensors:




•
13
Advanced cars have close to 100 sensors.
Smart homes use 10s and 100s of sensors.
Smart phones use now up to 18 sensors.
Medical diagnostics uses 10s of different sensors, which will be migrating to personal use.
It is thus not too big of a stretch to foresee the growth outlined by Bosch. Mobile Health
Health cost is dramatically increasing. •
•
In the US, it reached $2.5 trillion in 2009, representing 18% of the GDP (EE Times 12/9/09). Remote home care emerges as Tornado‐in‐making to reduce health care cost. Wearable (wireless) devices market (ABI Research) is forecasted to grow from 12M devices in 2010, to 420 million wearable health monitors in 2014.
•
•
• 59 million to be used at home. 500M people forecasted to be using mobile health apps by 2015 (www.alivecor.com)
2013 CES Show in Las Vegas:
• 19 Digital Health Sessions
• 350 exhibitors in Digital Health section
•
•
14
Medical Appcessories Sampler 1
Proteus digestible sensors send wireless signal through the body to a receiver. Records type of drug, the dose, and the place of manufacture.
Measures an heart rate, activity, and respiratory rate. 15
Low cost DNA chips containing up to 64 reactions of less than 1 μl volume.
Assay time is 10‐
30 minutes, cost < $1,000, assay cost $5‐$10 per chip.
Alcohol breathalyzer, $79 accessory.
Uchek (MIT) can detect up to 25 diseases, such as diabetes, urinary tract infections, and pre‐clampsia based cell phone camera reading.
It can also measure the levels of glucose, proteins, ketones, and more. Medical Appcessories Sampler 2
Ultrasound machine from Mobisante
went on sale in October 2011 http://www.mobisante.com/
LifeWatch V Tests ECG, body temperature, blood glucose, heart rate, oxygen saturation, body fat percentage and stress levels (heart rate variability). It contains health tests and apps, test result delivery, historical data analysis and even physician connectivity
http://www.lifewatchv.com/?page_id=480
16
EKG Monitor
http://www.alivecor.com/?gclid=CP‐
vwdDO4bUCFZOe4Aod4kcA8Q
Blood pressure monitoring cuff
http://www.deccanherald.com/content/230784
/monitoring‐your‐health‐mobile‐devices.html
Sensor based Appcessories Sampler 3
Augmented Reality
(Google Glasses)
Recon Goggles with heads-up display, GPS and IMU for skiers
Brainwave reading headset
http://www.ted.com/talks/tan_le_a_heads
et_that_reads_your_brainwaves.html
17
“Doctor in a pocket”
•
•
Scanadu is a 2010 startup based at NASA‐Ames Research Center.
Introduces in 2013 three home diagnostic tools based on imaging, sound analysis, molecular diagnostics, data analytics and a suite of algorithms to create a comprehensive, real‐time picture of your health. –
•
SCOUT: collects vital signs In less than 10 seconds contact with the temple:
–
–
–
–
–
–
•
Pulse transit time
Heart rate (pulse rate)
Electrical heart activity
Temperature
Heart rate variability
Blood oxygenation (pulse oximetry)
ScanaFlo: uses smartphone as a urine analysis reader. –
–
•
Uploadable to Scanadu smartphone app via Bluetooth.
Designed to be sold over‐the‐counter as a disposable cartridge, will test for pregnancy complications, preeclampisa, gestational diabetes, kidney failure and urinary tract infections. For pregnant women, is the first to provide a healthfeed throughout the duration of a pregnancy.
ScanaFlu
–
–
18
Assesses cold‐like symptoms, removing guess work from early diagnosis of upper respiratory infections. By testing saliva, the disposable cartridge will provide early detection for Strep A, Influenza A, Influenza B, Adenovirus and RSV. http://www.scanadu.com/news/#sthash.oeuESQSK.dpuf
Mobile Health
19
McKinsey: the Internet of Things (IoT)
 Internet of Things: sensors and actuators embedded in physical objects and linked through wired and wireless networks often using IP.
 Revolutionary aspects:
•
Deployment of physical information systems and ability to work without human intervention.
 In China’s 12th 5 year plan, IoT is one of the seven Strategic Emerging Industries with 5Billion RMB of government funding allocated during the next five years.
http://technode.com/2012/05/14/internet‐of‐things‐not‐just‐a‐concept‐for‐fund‐raising/
 Major Applications:
•
Information and analysis, such as
• Tracking behavior of persons, things and data through space and time,.
• Enhanced real time situational awareness of physical environment.
• Sensor driven decision analytics through deep analysis and data visualization.
•
Automation and control.
Source: McKinsey Quarterly 2010 Number 2
20
Market Segments for Internet of Things
21
Intel Context Computing
•
Sensors are key to the success of Context‐Aware Computing to sense e.g.:
•
All around me and my needs
• Understanding situations (e.g., mood of the person you meet).
•
•
•
•
•
•
Context‐aware algorithm development needs huge processing power.
•
•
All around devices
Personal health
Social interactions
Planet context
Universe context
100s of servers at HP.
Out of trillion sensors by 2020‐2022.
•
•
22
70% of sensors will be solving problems.
30% of sensors will create lifestyle enhancements.
Central Nervous System for the Earth
•
•
CeNSE is expected to deploy a trillion
micro/nano sensors and actuators by 2018.
Markets for CeNSE:
–
–
–
–
–
–
–
•
Climate monitoring
Oil exploration and production
Assets and supply chain tracking
Smart highway infrastructure
Tsunami and earthquake warning
Smart grid and homes
Structural health monitoring
Processing sensor information will require increasing the size of Internet 1000 times creating:
–
–
$70B global market for sensing systems, $290B market for value added sensing services.
http://www.hpl.hp.com/news/2009/oct‐dec/cense.html
Harbor Research performed a market research for HP resulting in CeNSE.
23
Trillion Sensors (TSensors) Vision
•
•
•
•
Mobile sensor market for volumes not envisioned by leading market research Trillion Sensor Visions
organizations in 2007, grew exponentially over 100,000,000,000,000
200%/y between 2007 and 2012. Several organizations presented their visions 10,000,000,000,000
for a continued growth to trillion(s).
Market research companies don’t yet see this 1,000,000,000,000
growth (see Yole’s forecast).
100,000,000,000
So the explosion to trillion(s) is likely to be driven by applications not yet envisioned by 10,000,000,000
"Abundance"
leading market research organization.
QCOM Swarm Lab, UCB
Bosch
Hewlett‐Packard
I launched TSensors Roadmap development* 1,000,000,000
Intel
to improve visibility of needed sensors to TI Internet devices
Yole MEMS Forecast, 2012
100,000,000
enable accelerated development.
TSensors Bryzek's Vision
Sensors/year
•
•
•
24
1st step: The TRILLION Sensor Universe, Conference at BSAC, March 6, 2013
2nd Step: TSensors Summit Conference at Stanford University with presentations by global sensor visionaries.
10,000,000
2007
10 year slope
Mobile Sensors Explosion
2012
2017
2022
2027
2032
2037
Why TSensors Needs Roadmap
•
Historically, sensor technologies have had long development cycles, about 20 years to volume production due to:
– Deployed “multi‐physics” and “multi‐bio‐chemistry” complexity.
– Lack of standardization (one product–one process–one ASIC‐one package‐one test system). •
•
•
Without focused commercialization efforts and availability of standardized development and manufacturing infrastructure, the development and scalability of sensors to trillions will be delayed.
Once visibility of likely ultra high volume applications emerges, focused commercialization efforts could be deployed.
Motivations for academia, Governments and industry to focus on accelerated sensor development:
– Acceleration of Abundance.
– Acceleration of new jobs creation.
– Accelerated market growth.
25
Cost Challenge
•
•
Global GDP will likely be around $130 trillion in 2013. Trillion sensors can’t represent more than perhaps 0.1% of GDP, implying the unit selling price under $0.13. –
•
Sensing element (or sensor arrays) with computational resources, communication, packaging and often powering capability. Such price level will create challenges for all involved technologies.
–
E.g., assuming a chip scale packaging, such smart sensing system‐on‐chip would have to be on the order of 1 mm2,
•
–
•
•
More than an order of magnitude footprint reduction from current state‐of‐art. At this die size, wafer requirement would be a challenging 100s millions of 12” MEMS/IC wafers.
Alternative manufacturing processes may be required to support required combination of volume and cost. Growth to trillions presents also significant scalability challenges to communication and data storage. –
26
E.g., Hewlett‐Packard was forecasting a need for 1000x growth of Internet needed by 2018 just to handle sensor data traffic. Global GDP is likely to reach $130 trillion by 2023. This forces smart sensor prices to a challenging $0.10 level. (Historical Data (blue) from Wikipedia. Extrapolation (red) by Bryzek.)
Digital Manufacturing to Rescue? •
3D printing comes in variety of flavors:
–
–
–
•
Kovio (San Jose) ships millions of RFID, NFC and EAS smart tags:
–
–
•
•
1000 transistors in 5 µm printed (30 µm thick) technology.
Potential for 2018: 10s of billions smart tags with sensors.
UC Berkeley is starting a program to print 0.2 µm transistors and sensors.
UMASS got $35M funding to develop roll‐to‐roll fabrication of transistors and sensors.
–
•
First printed home in 2014 in Amsterdam).
Printing toys, medical implants, etc.
Printing electronics.
Objective: reduction of equivalent wafer cost from $25,000 to $25 per square meter…
Digital manufacturing could be one of the enablers for fabricating trillions sensors at acceptable price.
27
Jobs, Jobs, Jobs…
•
Assuming an average revenue per employee in developed countries at:
– $200,000/year for component companies – $500,000/year (equal to 2011 average for the US NASDAQ 100 companies) for smart system companies. •
Assuming further:
– Average selling price of the sensor at $0.13 in trillion units/year volume. – Smart system selling price 50x higher* than combined sensors cost (current ratio).
•
Trillion sensors would thus represent in 2023:
– $130 billion component revenue level, translating to 650,000 new direct jobs.
– $6.5 trillion (5% of global GDP) smart system revenue, translating to 13 million new direct jobs.
•
Assuming indirect jobs multiplier of 3, this would result in 41 million additional jobs, for a total of 55 million jobs.
– Indirect job multiplier for knowledge workers has been between 2 and 4 (depending on region). – As a reference, the US created only 1.3M new jobs in the last 10 years, primarily in Government and medical sectors.
*Galvin G.: On the Road to $1T. COMS2012, June 2012
28
Jobs Where?
•
Sensor based systems require a high‐tech work force.
•
•
Majority of created jobs will likely be for knowledge workers. An example of a sensor based system could be Apple’s iPhone 4s, which had the following breakdown of 2011 selling price:
– 3% ($14) cost of assembly (China)
– 32% ($178) cost of components (global)
– 66% ($368) Apple’s share (US)
100% ($560) selling price
•
Most of “sensor and smart systems” jobs will likely be in industrialized countries.
Fortune Magazine 2011
29
TSensor Summit, Stanford, Oct 23‐25
•
TSensors Summit is organized as a forum for world sensor visionaries to present their views on:
– Application(s) and sensor types with a potential to generate the demand >1B/y by 2023. – Compelling justification why such high volumes are likely (e.g. overview of specific market drivers, global changes, problems solved, etc).
– Overview of current state‐of‐the‐art and needed technologies to support targeted sensor application growth.
– Characterization of major challenges in sensor development and commercialization.
– Estimated cost targets to trigger explosive growth for the proposed application(s).
– Proposed strategies to accelerate development and commercialization cycle.
• E.g., funding the academia‐industry‐Government consortiums, multi‐company joint developments such as cooptition (cooperating competition), etc.
– Estimated commercialization funding levels.
– Etc.
•
These visions will form a foundation for a published TSensors Roadmap.
30
Summary
•
Multiple organizations developed their own visions for trillion sensors applications.
– Leaving market research organizations behind.
•
•
•
•
Abundance is in synch with those visions.
The byproduct of sensor based systems growth to trillions could be the biggest bull market ever.
TSensors Roadmap emerges as an important factor to accelerate path to Abundance.
We need your support to help all of us on Earth to:
– Live longer and healthier
– In less polluted and more energy efficient world
– With more fun than ever.
31
Thank you