Underlying technologies for the. Internet of Smart Things

Underlying technologies for the Internet of Smart Things EPoSS Annual Forum 2009 Syracuse, Sicily, Italy 8-9th of October, 2009 Thomas J. Sommer Eur...
Author: Daniel Fields
4 downloads 0 Views 2MB Size
Underlying technologies for the Internet of Smart Things

EPoSS Annual Forum 2009 Syracuse, Sicily, Italy 8-9th of October, 2009

Thomas J. Sommer European Commission, Brussels DG Information Society and Media Unit G2 “Microsystems” Disclaimer: Any opinions expressed in this presentation are the author’s own and do not necessarily reflect the position of the European Commission

Outline of the presentation

• Smart miniaturised systems: the building stones of the Internet of Things • Ultra low-power requirements • RFIDs and OLAE • Potential IoT applications • Conclusion

••• 2

Smart environment

Drawing: courtesy of MIMOSA project ••• 3

What are Smart Systems? Smart Systems …intelligent miniaturised technical subsystems evolving from microsystems technology with additional functionalities:  are able to diagnose a situation, describe it and qualify it,  mutually address and identify each other,  are predictive,  are able to decide and help to decide,  enable the product to interact with the environment. They are networked, energy autonomous and highly reliable.

R&D in advanced microsystems driven by application

Smart implants

Smart RFID

Smart antenna

Courtesy of EPoSS

Smart tire

••• 4

Economic Opportunities and Social Impact



In 2006 the world market for microsystems technology: 40 billion USD projected to raise to 72 billion USD by the year 2011 (MST is ca. 25% of the Smart Systems market)



High growth potential particularly in the automotive, security and medical technologies sector, further in the consumer, communication and “portable” sector

Higher and strategic R&D investments in this area have an immediate impact on competitiveness and produce societal benefits Sources: YOLE Global MEMS/Microsystems, WTC Think Small , Gartner

Examples of technological challenges for the IoT

• Efficient power management for long-term autonomous operation of networked smart systems – Reducing of the leakage power (e.g. the ultra-low power NEM-FET switch), – Energy scavenging (different forms) – New battery types – UWB spectrum

• Reducing costs of RFID tags using OLAE • Inkjet printing of tags and WSN modules

••• 6

Challenges for power autonomous sensors

According to NEDO Roadmap: http://www.nedo.go.jp/

R&D on NEM-FET to reduce the leakage power • Currently running FP7 project NEMSIC • A NEMS device to co-exist with CMOS technology • The feature of the NEM-FET: suspended gate (SG-FET) • The NEM-FET (as a pre-transistor) will work as a sleeping transistor and will drive blocks of other transistors • Size: the goal: < 1µm (hundreds of nm) • Drastic reduction of the leakage current • NEM-FET switch expected to save >100 times stand-by power vs. MOSFET switch

Suspended Gate Field Effect Transistor (SG-FET) SG-FET Geometry

Device Level Simulations Proof of Concept 3D analysis of Suspended Gate Pull-in/Pull-out

Many parameters determine the SG-FET behaviour:  tox - the thickness of the gate oxide,  h - the thickness of the suspended gate,  Wbeam - the width of the beam, the length (L) of the device, Lbeam - the length of the beam, the width (W) of the device,  tgap0 - the gap between the oxide and the suspended gate,  kbeam - the lumped linear spring constant of the beam. Courtesy of the NEMSIC project, TU DELFT

Suspended Gate Field Effect Transistor (SG-FET) SG-FET vs FET (SPICE simulations) Conclusions: ION (mA/um)

Ioff (pA/um)

Delay(ps)

High Vt FET (nMOS)

0.51

1.6E+1

2E+1

SG-FET (n channel SG-FET)

0.50

8E-4

8.6E+3

• SG-FET is a viable alternative for FET as shared Sleep Transistor (ST) due to its extremely low Ioff. • Due to its rather large switching delay, this device appears not be suited for applications where the switching between active mode and sleep mode occurs too often. Area (um2)

32-bit ADDER

ION(mA)

Ioff (pA)

Delay ( us)

High Vt based ST

10

13E+3

SG-FET based ST approximation

10

7.2E-2

Total

Active

10

225

37

N/A (Related to Power Management Block)

Preliminary 900

136

Observations and Conclusions: • The delay for the original design is covering the entire “go to sleep”/”wake up” process thus we do not have information about the actual delay of the FET sleep transistor. • Ioff for SG-FET is 5 orders of magnitude lower then for High Vt based ST for the same ION and maximum circuit performance degradation of 5%. • The penalty to pay is maximum 4x increase in area of the ST Courtesy of the NEMSIC project, TU Delft

Energy harvesting (1) •

Energy harvesting technologies: – Mechanical, light, thermal, electromagnetic, from human body, other (chemical and biological sources) Min. 200 µW to be made available by an energy harvesting device

1. Mechanical Energy: electricity from vibration, stress, strain; forms of energy harvesting: electromagnetic, piezoelectric, electrostatic (capacitive) Example: EC project VIBES (finished in 2007) demo: energy aware wireless sensor node powered solely from vibrations; with 0.6 m/s² reading every 3 sec. 2. Light energy: - photovoltaic cells (silicon); - inexpensive flexible plastic solar cells (rollable, foldable); - long lifetime - power output level up to 10 mW/cm² - challenge: to conform to small surface areas

Courtesy Vodera Ltd. and Zartech Ltd, 2008

Energy harvesting (2) • 3. Thermal energy – Seebeck effect – Generated voltage and power: proportional to temperature difference – Rich source of energy in aviation: thermal gradient 50°C – Advantages: no moving parts, long life: 20 years, reliable – Drawbacks: low efficiency, large size But, Micropelt: nanostructured thermoelectric energy harvesters

• 4. Electromagnetic energy – Collecting parts of RF energy sources (radio, TV, mobile telephony) – Rectifying antenna: efficiencies above 90% – However: energy levels too low, not usable so far

• 5. Energy from the human body – E.g. kinetic energy of moving arm and heat flow from skin Courtesy Vodera Ltd. and Zartech Ltd, 2008

Efficient power storage Storing power during periods of inactivity:

• Rechargeable batteries – –

Used in consumer electronics products Li-ion: highest energy density, hundreds of charge/discharge cycles, no memory effect, holds the charge

• Thin-film batteries – -

polymer electrolytes, down to 50 µm; up to thousands of recharges, lose no power over time; Flexible, bendable; commercial production expected soon, e.g. Infinite Power Solutions, …

- Supercapacitors -

(electrochemical double layer capacitors)

Higher energy density than standard capacitors Long life Short charging time, no limits in numbers of chargings Low leakage No memory effects But, ageing of the electrolyte Courtesy Vodera Ltd. and Zartech Ltd, 2008

Cost issue • The global market: few billions tags per year • For item-level tagging the price of the EPC silicon transponder chip: major limitation • Significant price reduction expected if OLAE is used for RFID transponder chips • OLAE applied at low temperature, on low-cost substrates, at high processing speeds • OLAE RFID chips are thin and flexible (good for handling and yield) • OLAE is more adequate for integration of sensors into RFIDs • The challenge: to achieve OLAE tags with maximum adherence to the EPC protocol; important for acceptance • ORICLA (FP7 proposal under negotiations) addresses those issues

Source: ORICLA (proposal under negotiation)

Targets and Challenges • Passive RFID tags with OLAE technologies • HF-tag 13.56 MHz • Beyond the State-of-the-Art: – – – – –

Bi-directional communication, a.o. for ‘anti-collision’ High circuit speed: 25 kbit/sec Some work towards UHF range 860-960 MHz RFID tag at UHF with data storage of 4 bits Combination of novel n-type oxide transistors with p-type organic semiconductor technology - Processibility of oxide semiconductors at < 150 °C - Electron mobility > 0.5 cm²/Vs in an integrated process - Containing up to 3000 TFTs Source: ORICLA (proposal under negotiation)

RFID on paper

Courtesy of M. Tentzeris, GeorgiaTech

RFID on paper: characteristics

Courtesy of M. Tentzeris, GeorgiaTech

RFID/Sensor Module Integration Digital Data

Antenna Demodulation

Digital Logic & MODEM

Voltage Multiplier

Power

EEPROM

ADC

Modulation Digital data

Sensor

 Ultimate goal: All printed RFID tag (antenna, IC, battery, power scavenger and sensor) on paper  

Operating frequency: UHF (900 MHz), RF (2.45 GHz), scalable up to 60 GHz Suggested Module integration:

• Printed battery on surface • Printable sensor technology on surface • Surface mounted IC

Operation modes Passive Tags: – Antenna uses EM power from reader. Semi-Passive Tags: – IC uses EM power distribution – Sensor uses battery – Increased node’s lifetime

Courtesy of M. Tentzeris, GeorgiaTech

IoT applications (1) • Smart sensors and RFIDs to insert in critical structures (bridges, components of high speed trains, airplane wings, …)

to measure and signal the fatigue and wear of the material • Detection of cracks • At stake: saving of lives and money • Importance of ultra low power solutions; combination of power management, energy scavenging, batteries

IoT applications (2) Scenario: battery change / recharging for a Fully Electrical Vehicle (FEV) •

Goal: the smart system measuring the little remaining battery power in the car to signal this and the ID of the battery to the next battery exchange or recharge station.



How to achieve it? Using cyberspace to link physical world information to communities.



The system calculated the time available with the current battery.



Assumption: every FEV will have a GPS. The GPS determines the closest battery exchange station.



The database of the closest battery exchange station checks if a fully loaded battery would be available in the requested time. If not, a redirection or recharging is recommended.



IoT: battery exchange scenario for the fully electrical vehicle (FEV)

GPS satellite system

Ubiquitous ID resolution server

Product Info / Service server

4 2

GPS

uID

RFID/IoT communicator with R/W inside FEV 5

Ba tt.

Battery exchange station

3

1 FEV

Same type of battery ready for exchange

Battery with smart sensor and RFID tag in the FEV

Conclusion From Ambient Intelligence towards Internet of Smart Things

• Some of the technological issues for an affordable Internet of Smart Things (IoT) are NOT solved yet. • IoT is not only about governance, architecture, privacy and security. • IoT is very much also about middleware and technology. • Solving the technological challenges is the prerequisite for the vision of the future IoT.

For more information •

European research on the web:

• •

http://cordis.europa.eu/fp7/home_en.html http://ec.europa.eu/comm/research/future/



Information Society and Media:



http://cordis.europa.eu/fp7/ict/programme/home_en.html



Directorate G:



http://cordis.europa.eu/fp7/ict/programme/challenge3_en.html



Micro/nanosystems:



http://cordis.europa.eu/micro-nanosystems

Contact: [email protected]