Tesla User Interfaces

Wireless and Batteryless Tangible Interaction

Diploma Thesis, 2013

120 Pages, Grade: 1,3



1 Introduction
1.1 Motivation
1.2 Thesis Organization

2 Related Work
2.1 Mode of Operation
2.1.1 Basic Technology of RFID RFID Tags RFID Readers Reader-Tag Coupling
2.1.2 The 2013 RFID Taxonomy A Brief Overview Protocol Overview
2.2 RFID Interaction
2.2.1 Conventional RFID Interaction and derived Security Concerns
2.2.2 New Forms of RFID Interaction The Work of Nicolai Marquardt et al. The Work of Joseph Paradiso et al.
2.2.3 Sensor Enriched RFID

3 Interactive RFID Prototypes
3.1 Paper RFID Tags
3.1.1 Components Basic Paper RFID Components The SM130 RFID Reader
3.1.2 Paper RFID Tag Types Simple Tag Poti Tag Slider Tag Tilt Tag
3.1.3 Summary of the Paper RFID Tags
3.2.1 AVRFID Components 125 kHz Reader The ATtiny85 Microcontroller
3.2.2 Programming AVRs Implementation of the EM4102 Protocol Programming the ATtiny85
3.2.3 AVRFID Rating
3.3 Antennas
3.3.1 General Antenna Knowledge Independent Variables for Antennas in Near-Field Coupling Independent Variables for Antennas in Far-Field Coupling Antenna Tunning
3.3.2 Practicing Antenna Design
3.4 Comparative Summary of the Prototyping Approaches

4 Tesla User Interface
4.1 Hardware
4.1.1 Elaborated SM130 Reader Setup
4.1.2 Liberated Control Widgets Button Control Widget Rotary Control Widget Slider Control Widget
4.2 Software
4.2.1 Arduino Firmware
4.2.2 Processing Software
4.3 Toolkit for Designers

5 Implications for User Interaction
5.1 Design Space Definition
5.2 Fields of Application
5.3 Ways of Interacting

6 Discussion
6.1 Limitations
6.1.1 Technological Barriers
6.1.2 Within Interaction
6.2 Future Directions
6.2.1 Future Work
6.2.2 Future Trends

7 Conclusion


List of Figures

List of Tables
A Collection of RFID Scenarios
A.1 Day in the Life
A.2 Airport
A.3 Wild List of Collected Scenarios
B Miscellaneous Addendum
B.1 Source Code
B.1.1 Arduino Firmware Versions
B.1.1.1 Arduino Firmware for a RDM630 Reader Module
B.1.1.2 Arduino Firmware for the elaborated SM130 Reader Setup
B.1.2 Assembler Code for implementing the EM4100 protocol
B.1.2.1Original Assembler Code
B.1.2.2 Additional Assembler Code
B.2 EM4102 Assembler Parity Calculation Example
B.2.1 Row Parity Calculation
B.2.2 Column Parity Calculation:
B.3 Additional Pictures and Schemes
B.3.1 SM130 RFID Reader Module
B.3.2 RDM630 RFID Reader Module
B.3.3 Sparkfun AVR Pocket Programmer
B.3.4 Small wired 13.56 MHz Tag
B.3.5 Scheme of the Button Control Widget
B.3.6 Scheme of the Rotary Control Widget
B.3.7 Atmel ATtiny85 Block Diagram


"The economic transmission of power without wires is of all-surpassing importance to man. By its means he will gain complete mastery of the air, the sea and the desert. It will enable him to dispense with the necessity of mining, pumping, transporting and burning fuel, and so do away with innumerable causes of sinful waste. By its means, he will obtain at any place and in any desired amount, the energy of remote waterfalls — to drive his machinery, to construct his canals, tunnels and highways, to manufacture the materials of his want, his clothing and food, to heat and light his home — year in, year out, ever and ever, by day and by night. It will make the living glorious sun his obedient, toiling slave. It will bring peace and harmony on earth." — Nikola Tesla, 1905.[77]

1.1 Motivation

For the exploration of new technological possibilities, it is indispensable to understand what technology is actually to our hands in that specific moment. Every once in a while the time of technological inventions was just not ripe enough to unfold its full potential. Radio frequency identifiaction (RFID) undoubtedly is such an example. RFID maybe still takes some more time to mature its potential to a full extend, but surprisingly innovative and unused possibilities are ready to be used in creative ways, as this ongoing thesis will show.

Having already found its way into every day lives since 30 years, people rarely recognize RFID, nor its dozing potential for new ways of user interaction with their environment. The scientific journey, that was needed to achieve such a level of appliance for this tech- nology, almost spans two centuries. Beside Nikola Tesla, numerous well-known scientist contributed a remarkable share. One scientific cornerstone for this work lays in the dis- covery of the electromagnetic induction by Michael Faraday in the year 1831. Together with Tesla’s polyphase alternating current inventions and a fundamental theory work by James C. Maxwell, Rudolf Hertz must be credited to be "the first to ever transmit and receive radio waves"[36]in the year 1887. Another 50 years later, Harry Stockman wrote an article called "communication by means of reflected power"[73]and finally this sequence of inventions could slowly be called radio frequency identification (RFID). The first com- pletely passive RFID versions emerged in the 1970s, especially with the work by Koelle et al.[32].[36]

Today, in a time rapidly heading towards the peak oil1 and facing an energy crisis, governments all around the world are forced to search for alternative power supplies. But it seems, that another solely replacement for oil as an energy source, is not likely to be found. If anything will work, then it is only a holistic approach that saves every little energy consumption possible. At the same time, already living in a post PC era, in which every person owns multiple electronic devices, Mark Weiser’s ubiquitous spirit is as obvious as it is within Tesla’s words (above). Further it is undeniable, that the demand for increased networking will not lessen in the foreseeable future.

The primary advantage of standard RFID tags was initially that they do not require a line of sight, as compared to how the graphical bar code opponents are scanned, for example. But production costs could not be compensated by the respective business models for a long time. That’s why the degree of adoption is not as large as it possibly could. Roy Want states that it is just missing "the right technology catalyst to kick- start the ecosystem"[81]. But lately, wide spread efforts are undertaken to add extra functionalities, which start to pay off for RFID.2 Exactly this conjuncture of ubiquitous computing [82]trends, the energy crisis and the enhanced functionalities, will be the big chance for RFID to implement its full capabilities right now. Therefore I am convinced, that the self-sustaining nature of completely passive RFID tags is one of the goal key advantages, which lets them be the tool to realize this ubiquitary dream of the internet of things.[79]What this is virtually implicating for ways of prototyping and designing future user interfaces and how accurate user interactions can be applied to these trends, will be one of the big research questions of this thesis.

In order to gain more knowledge about the implementation details of the RFID technology, this thesis evaluates in a top-down approach the complete status quo of RFID at the present state. After identifying practical background knowledge for innovative RFID system designs that will also work in practice, this technology gets reassembled in a bottom-up style. That way it is guaranteed that all important aspects are taken into account for the proposed development of higher level user interaction concepts. Furthermore an overview helps to clarify concrete topics for the future work in this field.

1.2 Thesis Organization

Related Work : Because of the interdisciplinary complexity, this chapter tries to provide a detailed but compact step into the field of RFID. The first section deals with the different modes of operation and the 2013 RFID Taxonomy. Afterwards the focus shifts to the interaction possibilities with the RFID technology, especially looking at the new forms that evolved in related research. This chapter ends with a glimpse into the field of sensor networks, pinpointing promising combinations with RFID technology for the realization of i.a. wireless and batteryless physical control elements.

Interactive RFID Prototypes : An extensive work of this thesis, was the building of sev- eral interactive prototypes, that all rely on different RFID technologies. This also means, accomplishing multiple entry tunnels, in order to achieve a state of reliable operation. The goal of building these prototypes is the evaluation of their technical benefits and limitations.

1.2 Thesis Organization

Tesla User Interface : This section presents the concept of a Tesla User Interface. An interface whose controls work wirelessly and without the need for batteries as it completely relies on the passive RFID technology presented before. By that a boost in the design flexibility of physical user interfaces can be achieved.

Implications for User Interaction : Considering the three series of prototypes, various implications were observed. In the first part of this section the influential properties and their according implications for the design space are presented. This yields into an ordered structure for the fields of application and a few scenarios are illustrated. The last part of this chapter sums up all identified concepts, that in certain way can be applied for applications using the RFID technology.

Discussion : Utilizing the RFID technology for innovative ways of user interaction definitely demands for discussion in multiple perspectives. A summary of technological barriers of RFID mark the beginning of the identified limitations. Then this leads to the boundaries that derive for concepts of user interaction relying on this technology. In order to face these constraints for science in this field, a detailed future work describes the issues that should be tackled next. Therefore this is structured according to the affected research subjects.

Conclusion : The most important results of this technical evaluation of RFID for the appliance of innovative user interactions are concluded in this part. Therefore it relates to parts of the future work and suggests ways to handle the RFID technology in future HCI project.

Appendix A - Collection of Scenarios : It takes holistic expert knowledge and a huge breeze of creativity to develop scenarios that exploit the ubiquitous possibilities with RFID to its full potential. In order to draw a complete picture of possible fields of applications, a huge collection of scenarios is listed in this part of the appendix.

Appendix B - Miscellaneous Addendum : Appendix B provides the different firmware source codes that were presented and an example parity calculation of the EM4102 protocol is illustrated in detail. A few yet unpublished circuit schemes are closing this thesis.

Related Work

"At first glance, the concept of RFID and its application seems simple and straightforward. But in reality, the contrary is true. RFID is a technology that spans systems engineer- ing, software development, circuit theory, antenna theory, radio propagation, microwave techniques, receiver design, integrated circuit design, encryption, materials technology, mechanical design, and network engineering, to mention a few"[36]. This was part of Jeremy Landt’s conclusion in an article about the history of RFID, that gives a first hon- est impression of the diversified technology that is RFID. Therefore this chapter tries to provide a detailed but compact step into the field of RFID. The first section deals with the different modes of operation and the 2013 RFID Taxonomy. Afterwards the focus shifts to the interaction possibilities with the RFID technology, especially looking at the new forms that evolved in related research. This chapter ends with a glimpse into the field of sensor networks, pinpointing promising combinations with RFID technology for the realization of i.a. wireless and batteryless physical control elements.

2.1 Mode of Operation

In this section the radio frequency technology is explained by having a closer look at the inner workings of RFID tags, RFID readers, their low-frequency/near-field or high- freqeuncy/far-field coupling techniques and other important parts of the RFID communi- cation. Additionally the 2013 RFID taxonomy, which came out as a side product during the research for this thesis, will be briefly presented, to complete the presentation of RFID technology.

2.1.1 Basic Technology of RFID

In general the communication of RFID is composed of two interacting parts - the RFID reader and the RFID tag. First of all the various characteristics of RFID tags are explained. Then the role and the RFID reader’s modes of operation are explained. Afterwards the interaction of this communication technology gets outlined.

illustration not visible in this excerpt

Figure 2.1: RFID Tag Properties RFID Tags

There are probably as many different forms of RFID tags as there are existing fields of application for this technology. Breaking down all the possible combinations of properties (see figure 2.1) these tags or transponders can have, is essential to fully understand RFIDs. It further helps to identify less implemented variations and most importantly new ones with potentially interesting combinations.

The general construction of an RFID tag consists of three components. First of all there is the integrated circuit. That is a microcontroller with a certain identification number according to the specific protocol that is used for the communication with the reader. The microcontroller itself consists of a memory, which can have read-only, write-once- read-many (WORM) or read/write characteristics. Furthermore it has certain electrically erasable programmable read-only memory (EEPROM), random-access-memory (RAM), static random-access memory (SRAM) memory specifications and can usually save user data of around 200-8000 Bit[80]. But already in 2006 Hewlett-Packard came up with a 4 megabit solution1, which would enable storing of audio or video files on smart posters. Further some RFID chips also provide the possibility to assign a password protection to read/write operations. Thanks to Moore ’ s Law2 the size of these microcontrollers will continue to decrease hopefully. One example into this direction is the 0.05 x 0.05 mm RFID powder chip which Hitachi presented in the year 2007.[76]This on its own opens the door for various science fiction scenarios, which may not be so far away anymore (more details in Appendix A - Collection of RFID Scenarios). Sometimes certain RFID chips have integrated components like capacitors or rectifiers, which influence the tags’ perfor- mance noticeably (see figure 2.2). Therefore anyone who is prototyping or developing RFID tags should understand the electronic dependencies with these components. Hence, a main challenge will always be to tune an RFID tag circuit’s components for ideal per- formance with a certain RFID reader. This also means, being highly aware of the power consumptions and operating voltages.

illustration not visible in this excerpt

Figure 2.2: Exemplary RFID Tag Block Diagram of the EM4100’s Integrated Circuit[65]

Another important point to mention is the existence of a few more sophisticated RFID tags, which are able to operate at two or more different frequencies. Besides this, there is a rising variety of enhanced RFID tags, incorporating passive sensors (more details in

2.2.3 Sensor Enriched RFID). In this context also extra alternative power supplies[74]and an continuous data storage of measured values are of interest. Power supply probably is the strongest distinctive feature that RFID tags have. They are classified into four types: passive, semi-passive, semi-active and active RFID tags.[4, 29]

A completely passive RFID tag only has that kind of current available which is induced when a tag’s antenna enters the electromagnetic field of a reader. Due to the simple circuit design of passive RFID tags, they can be produced for very cheap prices. On the other hand they only have very limited read ranges depending of the underlying near-field or far-field setup.

The semi-passive RFID tags differ from the passive ones in that way that they have an additional battery power supply for the integrated circuit and only for it. That means that the response to the reader still relies on changing of the electromagnetic field like it is done by completely passive RFID tags, with the benefit of little farther read ranges, because there is no loss of energy for powering the IC. Sometimes they are also called battery assistant passive (BAP) RFID tags.

In an active RFID tag the IC is powered by a battery. Most of the times it period- ically transmits its ID in 1 to 15 seconds intervals, also known as the beacon rate. This results in bigger read ranges (far-field) with the disadvantage of bigger tag di- mensions, higher production costs and a dependency on battery lifetime according to the number of reads.

The semi-active RFID tag is almost the same as an active RFID tag. The difference is, that it only sends its ID, when interrogated by an RFID reader. This improves battery lifetime appreciably.

From the view of electronic engineers and computer scientists the most challenging type of RFID certainly is the passive tag, which I therefore want to concentrate on in this ongoing thesis. These self-sufficient devices have the highest potential for ubiquitous com- puting. As already mentioned there are huge price differences depending on the individual properties a tag is equipped with, which depend on the respective field of application and on the amount of produced tags. Despite the various antenna shapes[72]like e. g. dipole or loop, another characteristic, which also depends on the field of application, is the actual appearance of the tag. One of the biggest advantages of RFID tags is that they don’t need a line of sight in order to operate. Passive RFID tags additionally have the benefit that they can be completely sealed in plastic, rubber, concrete, glass. Hence almost any mate- rial can be used, as long as the radio frequency communication is not weakened (like with conductive/metallic materials). Independent of ethical considerations, this also includes implants into animals and humans. Furthermore it can easily be embedded into stickers or garments just to name a few. RFID Readers

illustration not visible in this excerpt

Figure 2.3: RFID Reader Properties

Unlike RFID tags, RFID readers will always have a fixed or at least an adequate power supply, as it is supposed to power the RFID tags as well. Sometimes it is simply called interrogator, because it is always supposed to be the one that enables the communication. Currently, the amount of RFID readers, implementing more than a single protocol are slowly rising. But most often, and especially when prototyping, one still has RFID reader modules, that only support one specific protocol. This is closely coupled to the number of frequencies a RFID reader can support. At a certain point this also means that a RFID reader should not only be able to switch between different protocols and frequencies. But in order to do so, it also has to be able to switch between different antenna types, like e. g. dipole and loop antenna designs[72]. Concerning the frequency, a few new readers support frequency hopping spread spectrum (FHSS)3, as a kind of automatic tuning technique. It searches for the frequency in the communication with a RFID tag, which yields in a better performance. In the slow hopping variant at least one bit is transmitted until the frequency hops, and in the fast hopping variant the frequency could hop one or more times while transmitting a single bit.

2.1 Mode of Operation

Besides the antenna type, RFID readers can have linear or circular antenna polarisation configurations. An antenna can be connected via one or multiple ports and could be inte- grated with the other components onto a PCB board etc.. But most often it is connected in an external set up. In this case there is the possibility to connect multiple antennas to a single reader. In consequence the antennas may either have different types in order to support different frequencies, or they are of the same type. If the latter is the case the outcome of this is either a parallel or an orthogonal[80]configuration of the multiple antennas.

The microcontroller on the RFID reader module implements the reader side of the communication according to the respective RFID protocol on an integrated circuit (IC). Most RFID readers are only capable of one protocol, even though there exists a slowly rising number of RFID readers, supporting multiple protocol standards4. As we assume a quasi unlimited power supply of the reader, it is not a critical design component. Except for one thing, which could be part of the RFID IC of the reader. Depending on the radio frequency (RF) coupling of tag and reader, it is either important how sensitive the radio receiver is or how accurate the flow of current measurement works (more details in the next paragraph Reader-Tag Coupling).

The RFID reader module can have different communication interfaces. The serial stan- dard RS-232 is a common way to receive read RFID tag data on a computer. I used a RFID module connected to an Arduino board, which in turn is connected via USB to the PC5. The software on the computer therefore only has to implement the serial library in the according programming language (e. g. Processing). The difference between the RS-232 and the RS-485 standard is that the RS-485 is not ground related and has a symmetrical conductor. Therefore it is not lost to common-mode interference. The latest RFID readers currently available on the market are often directly equipped with USB6, Ethernet7 or even wireless LAN8. Further RFID readers have the obligatory feature of either being placed stationary to a fixed position or coming in a mobile handheld variant e. g. as a personal digital assistant (PDA) or smart phone with NFC functionality. Reader-Tag Coupling

One reason of the huge diversity of RFID tag properties are the two main ways of coupling between a reader and a tag resulting in a better near-field or far-field communication. Explanations of the technological method of operation of the reader-tag coupling in the related literature are often unclear and imprecise.[74]That is why in the following both ways will be explained in depth, following the definitions of Roy Want [79, 80, 81]. Related to the many tag and reader properties, as described in the previous sections, there is a diverse spectrum of frequency bands, which can be deployed for RFID applications. This underlies the result of a combination of issues. First of all, not every frequency is suitable for every communication task, as the electromagnetically characteristics of wave length λ and photon energy E enable different electronic possibilities. Another issue are the country specific regulations for frequencies and their usage. Especially Europe, Asia and the USA do have different liberated frequency bands for the same fields of application. In figure 2.6 an overview of the frequency bands for RFID communication can be seen. It is made up out of the low frequency (LF) band with frequencies between 30-300 kHz, the high frequency (HF) band with frequencies between 3-30 MHz, the ultra high frequency (UHF) band with frequencies between 0.3-3 GHz and a very few RFID applications even using the super high frequency (SHF) band with frequencies over 3 GHz.

illustration not visible in this excerpt

Figure 2.4: Near-Field Low Frequency RFID Coupling Set Up. Illustration by Roy Want [79]

Near-Field The near-field communication exclusively uses the low and high frequencies. In the low frequency band 125 kHz is the most popular frequency and in the high frequency band it is 13.56 MHz, which is also used by the NFC standard. A RFID reader and a passive tag interact within a range of usually under 1 m. In figure 2.4 the powering and the data transmittance on the basis of resonant inductive coupling is illustrated and will be explained in the following.

The reader passes an alternating current through its antenna coil producing an alternat- ing magnetic field around it. If a RFID tag with its smaller antenna coil enters this field, an alternating voltage will appear across its coil ends, similar to how a power transformer works. Now this voltage is rectified and coupled to a capacitor, accumulating a reservoir of charge. At a certain threshold of stored load the tag’s integrated circuit gets sufficient power and will start working. In order to transmit data back to the reader, a tag applies a small current to its own antenna coil, resulting in its own small magnetic field. This modifies the readers magnetic field in such a way, that the reader can detect a tiny vari- ation in current flowing through its coil. These variations in current are proportional to the load applied by the tag to its antenna coil. That is why this form of coupling is called load modulation and not back scattering, as often commonly misinterpreted. This means that a tag can encode a signal simply by varying its load appliance to its antenna coil for representing its ID according to the respectively declared protocol. On the other side the reader monitors the flow of current in its antenna coil to then recover the signal transmit- ted by the tag.[80]Thus a critical parameter for RFID reader designs for the near-field communication is the accuracy of the reader monitoring and detecting variations in the flow of current in its antenna coil. The more sensitive and precise, the bigger the read distances will get. Besides this, the operational range of magnetic induction is determined by the equation

illustration not visible in this excerpt

with c being the constant for speed of light, π being the circle constant and f the frequency. This means that with an increasing frequency the operational range for magnetic induction decreases. That is why magnetic induction for RFID communication is limited to the near-field range and is only used from low (LF) to high frequencies (HF). Additionally the available energy for induction is a function of the distance to the reader antenna coil. "The magnetic field drops off at a factor of

illustration not visible in this excerpt

is the separation of the tag and reader, along a center line perpendicular to the coil’s plane. So, as applications require more ID bits as well as discrimination between multiple tags in the same locality for a fixed read time, each tag requires a higher data rate and thus a higher operating frequency. These design pressures have led to new passive RFID designs based on far-field communication."[80]

illustration not visible in this excerpt

Figure 2.5: Far-Field High Frequency RFID Coupling Set Up. Illustration by Roy Want [79]

Far-Field High-frequency RFID systems using the ultra high (UHF) and super high frequency (SHF) bands are equipped with dipole antennas. In the UHF band 869-920 MHz9 and 2,4 GHz frequencies are most often implemented. In order to absorb maximum energy out of the electromagnetic waves, propagated by the reader, the length of these smaller tag antennas has to be extremely well tuned to the respectively used frequency. The electromagnetic wave capture induces "an alternating potential difference [...] across the arms of the dipole. A diode can rectify this potential and link it to a capacitor, which will result in an accumulation of energy in order to power its electronics. However, unlike the inductive designs, the tags are beyond the range of the reader’s near field, and information can’t be transmitted back to the reader using load modulation. The technique designers use for commercial far-field RFID tags is back scattering"[80]. As the tags’ integrated circuit gathered enough power to start working, it changes its impedance over time, encoding the tag’s ID according to the respective protocol. "In order to change the impedance of the antenna, a transistor is placed between the two branches of the dipole antenna. When the transistor conducts current, it short circuits the two branches of the antenna together, changing the antenna impedance"[66]. Every time an impedance mismatch occurs at the tag’s dipole antenna, some of the incoming wave energy is reflected back towards the reader. At the reader a sensitive radio receiver detects the tiny waves of energy reflected by the tag. The sensitivity of the reader’s radio receiver is an essential part in this high frequency system (see figure 2.5). "The actual return signal is very small, because it’s the result of two attenuations, each based on an inverse square law—the first attenuation occurs as EM waves radiate from the reader to the tag, and the second when reflected waves travel back from the tag to the reader"[80]. So the odd wave energy recurring to the reader is therefore determined by the separational range r.

illustration not visible in this excerpt

Let us remind of another physical coherency. The higher the frequency the more directional the radiancy spreads in space, similar on how higher acoustic frequencies behave. This more focused energy wave is another physical aspect why in this context higher frequencies support bigger read ranges.

So far, RFID communication involves the different frequency bands and the two radio frequency coupling techniques. In figure 2.6 an overview of other important aspects of the communication can be seen and I briefly want to explain them in the following.

The most common form of communicative relation in RFID, is a point-to-point relation, happening with one RFID reader interrogating one single RFID tag at a certain time. Not to be confused with the fact that there is normally only one RFID reader responsible for reading a greater amount of RFID tags in its life time. Nonetheless more advanced RFID systems exist, which provide a multipoint, multicast or broadcast type of relation. There- fore some are capable of bulk RFID tag readings, because its protocol supports a form of media access control (MAC) for the RFID tags, also known as an anti-collision mechanism. Almost all of the existing mechanisms work on the basis of a time devision multiple access (TDMA) algorithm, which is either tag-driven (asynchronous) using Aloha techniques or reader-driven (synchronous) using tree algorithms.[68]Theoretically, the other forms of division multiple access, e. g. space, frequency and code could still be potentially useful in the context of anti-collision mechanisms. A space division multiple access (SDMA) method could make use of multiple readers separated in space or placed in an orthogonal orientation to each other. With the rising number of RFID tags and readers,working with multiple protocols in multiple frequencies, even something like frequency division multiple access (FDMA) could get possible. So far code division multiple access (CDMA) could theoretically only be interesting for separating different RFID reader signals within a tag, letting the tag chose, which reader it wants to communicate with. For the sake of completeness I only want to mention that collisions could not only occur between tags while being read, but also between overlapping RFID reader fields (more details in my 2013 RFID Taxonomy mind map).

illustration not visible in this excerpt

Figure 2.6: RFID Communication Characteristics

The flow of data in an RFID communication so far works in halfduplex (hdx) most of the times, because either one side transmits its data at a certain point of time. Little academic informations can be found on fullduplex (fdx) or sequential (seq) modes of data flow[19], but in code of practice there is no protocol implementing them so far. There are several encoding formats for the data itself like e. g. the Manchester Biphase-L10 or the Manchester IEEE802.3 definition, the Wiegand format and Phase Shift Keying (PSK) or Frequency Shift Keying (FSK).

Another important issue of the communication is the security of the communication channel. The limited computational resources of the RFID tags make it difficult to implement elaborated encryption techniques like it is possible for conventional wireless information transmittance. If a cryptographic security mechanism is implemented, a symmetric or asymmetric challenge-response authentification (CRA) is most likely used. The asymmetric CRA is also called public key cryptography and works with e. g. pseudo number generators (more details in 2013 RFID Taxonomy mind map).

2.1.2 The 2013 RFID Taxonomy

This section gives a brief overview of the 2013 RFID Taxonomy as a whole and ends with a closer look in the diverse spectrum of existing standards and protocols. A Brief Overview

During the process of sorting the related work and getting an overview about RFID and its forms of interaction, the 2013 RFID Taxonomy in figure 2.7 came out as a side-product. I think everyone, who attends to the RFID technology, will sooner or later arrive at a confusing point with datasheets, that are hundreds of pages long and corporations claim- ing their product to be the most successful and most used one. This and the immense

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.7: An overview of the 2013 RFID Taxonomy

heterogeneity and incompatibility thwarted the development of interesting new RFID ap- plications a lot. So in order to give other scientists and designers a quick and honest overview about the different fields and technologies that RFID is today, I published the original Mind Map project on the blog of this thesis, and anyone is free to contribute to it in order to keep it complete and up-to-date11. I hope this to be a valuable contribution, because only few scientists stepped out of their somewhat exotic research fields, classifying their work accurately and by that providing an holistic understanding of this technology. In some parts, existing taxonomies [26, 68, 83, 42] and the basic structure by Le et al.[39] found their way into my taxonomy. It is meant to complement the outdated parts and adds new categories, in order to cope with the many aspects and dependencies of RFID in terms of ubiquitous computing and designing new user interactions. The basic structure by Le et al. corresponds to an own ISO/OSI layer system for RFID, which is maintained in my RFID taxonomy version, even though it had a strong business process point of view to it.12 Similar to the well known TCP/IP protocol stack, which builds the foundation of the internet, it is desirable to have such a protocol layer system for the different demands of an internet of things with RFID as well (see figure 2.8 Proposed ISO/OSI-Layer System by Le et al. [39]). The RFID ISO system starts with the physical layer, representing all the low-level characteristics of RFID readers and tags (see figures 2.3 and 2.1 for an extended view) being necessary for the different variations in communicating with each other (cf. figure 2.6). On top of the communication layer a middleware adds a service integration and delegates relevant RFID information to the respective application in the application layer.

Figure 2.8: Proposed ISO/OSI-Layer System by Le et al.[39]

Building upon those four basic categories, I introduce new ones, which meet other im- portant fields of RFID, especially for developing new applications and forms of interaction (see figure 2.7). The development category shows ways of prototyping with the RFID tech- nology and how proven techniques of evaluation and production are summed up. Hand in hand with the evaluation techniques, the user category lists reviewed ways of letting people interact with the RFID technology and further gives a collection of possible stake- holder roles. Like in any other information system, security is an important aspect, so that an own category represents best practices for secure RFID interaction and gives an overview about the types of potential risks. The only category, that is not formulated in a general representation of knowledge, is the protocol category, which provides an practi- cal overview of the vivid scenery of proprietary corporate protocols, the official ISO/IEC standards and everything in between. Unfortunately a more detailed explanation of the 2013 RFID Taxonomy would go beyond the scope of this thesis right now13. Protocol Overview

The subject of RFID is affected by a vast heterogeneity of corporate protocols that itself rely on one or more official ISO/IEC standards. The multiplicity of standards and their different combinations make RFID a well-founded black box technology for most people. The ISO standardization categories are seperated in air interface protocols, data content, conformance and application protocols. Further there exist ISO standards for nearly each field of application for RFID14. An important player in the standardization process is GS1, which manage the EPCglobal protocols15 and is among other companies a sponsor of the Auto-ID Lab16. Further there a more specific organisations like the Zigbee Alliance17 and a few Open Source projects like the OpenPCD and Fosstrak protocols (both working at the 13.56 MHz frequency), as well as the Open Source middleware solutions AspireRFID, Rifidi Edge Server and Accada. Besides that, a great variety of corporate protocol devel- opments exist, which partly rely on the mentioned official standards. Among the most important companies there is EM Microelectronics known for their EM4100 protocol18 and many others for the LF, HF and UHF frequency bands. Further there is Atmel with a couple of protocols in the LF frequency band (e. g. known for T5557 protocol19 ) and HID Global20 that have passed a few protocols. Last but not least there is NXP Semiconductors that developed the widespread Mifare protocols21 and together with Sony and Nokia have lately forced the near field communication (NFC) standard that also works at 13.56 MHz.

2.2 RFID Interaction

After having discussed all the technological aspects of RFID, this section concentrates on how people use this technology today and what kind of new forms of interacting are proposed by the literature. First common usage of RFID is described. Followed by listing new interaction approaches, by referring to two of the most important related researchers: Nicolai Marquardt and Joseph Paradiso. The last part takes a closer look on the integration of RFID with wireless sensor nodes (WSN), in order to understand the potential benefits for future interactions.

2.2.1 Conventional RFID Interaction and derived Security Concerns

Despite the many fields of application that not happen to have humans interacting directly with RFID (because it mostly enables automated mechanisms like e. g. in logistics and supply chain management) most people have had contact with RFID in the matter of identification. This includes passports, opening doors and time logging. Furthermore many people know RFID from paying their food in the cafeteria, paying for their car with automated toll collection22 or paying for public transportation23. All these adoptions, except the toll collection example, assume a user to hold his card close to the reader device, so that a back-end system can check an ID on what actions to initiate or allow. Therefore the user relies on feedback by the reader device to understand the state of the system (cf. The Work of Nicolai Marquardt et al.). This traditional system consists of the people using RFID tags, being read by RFID readers, which in turn are connected via a middleware, delegating the information to a main software and checks the ID within

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.9: Privacy and Security Threats in RFID Applications by Garfinkel et al.[21]

a database. The complexity of this system is often criticized to expose severe security risks. In figure 2.9 the illustration by Garfinkel et al.[21]identifies all affected elements in the system with their according type of privacy or security risks. In order to explore and face all kinds of privacy and security risk, Rieback et al.[63]performed valuable work in this field. An example is an electronic device called RFID Guardian, which manages the rights to read or block ones own surrounding RFID tags.

2.2.2 New Forms of RFID Interaction

In a world full of ubiquitous distributed RFID sensors, another important aspect also is how these informations are read and bundled. This is primarily determined by the out- come of reader devices in terms of look and feel and how they work. Besides the fact that popular mobile devices are more and more equipped with NFC capabilities24, and there- fore certainly will become an important tool in accessing sensor networks, there are also a few other approaches. Two of them are presented by Joshua Smith et al.[70]. They have experimented in detecting tagged objects using their so called iGlove and iBraclet devices. The iGlove is a pair of gloves equipped with reader units. Tests in a medical laboratory showed that it was possible to track objects and retrace users interaction with them[58]. Although "the early prototype was too crude for true long-term deployment"[70], the form factor seemed acceptable for this special medical exertion. The iBraclet opens this up to a broader user population, because it is "aesthetically and ergonomically much [more] pref- ered to gloves. [...] However, the wearable-reader approach still involves open questions about basic feasibility. It is not possible to know in advance what combination of size, aesthetics, and battery life will satisfy a picky consumer or if such a satisfactory combina- tion even exists"[70]. Simliar to the iBraclet, Berlin et al.[6]have built a wrist worn reader device, that also detects the grasped objects. They have especially put effort into antenna tunning and improving read ranges. Further they have developed an initial, practical and fast benchmark study test for these kind of user interactions. It is performed by simply unpacking and repacking a box with tagged objects by the study participants. Hence it is called The Box Test.[6]Overall it is mentionable, that in the field of human-activity detection there are three main techniques: computer vision, active sensor beacons[75]and passive RFID.[70]

An example of interaction with a fixed installed RFID reader is proposed with the TangiSense table by Kubicki et al.[35]. The table consist of a grid of RFID reader antennas, sensing the objects standing upon it with an algorithm of space division. "The interaction is no longer the result of touching the table, but comes from handling tangible objects placed on the table. The use is closer to the natural use of a table.[...] Starting from this principle, the article proposes the use of RFID tags to collect the elements of context awareness in order to adapt the workspaces to the various possible situations around a table (work alone or with several users, on a common or individual space)"[34]. A few works exist using such a table for gaming purposes [41, 44]. The Work of Nicolai Marquardt et al.

Nicolai Marquardt, Alex S. Taylor, Nicolas Villar and Saul Greenberg consequently identi- fied problems in the field of RFID interaction and deduced simple enhancements for RFID tags from it. First of all they separate between tags with no (or little) personal informa- tion e. g. transit passes or concert tickets, and tags with personal data. These are further classified into ones containing personal informations, which are not that fundamental like e. g. customer loyalty cards, and the ones storing personal data, which represent high-risk privacy sensitive data like in the trend of recent RFID enhanced credit cards or passports. The identified threats include: unauthorized scanning, unauthorized location tracking of individuals, eavesdropping of authorized communication, leakage of biometric data stored on RFID tags, hacked RFID deployments and cloning of cards (e.g., [27, 33, 38]). The severity of these threats stand to reason, when regarding the vastly incomplete or wrong mental models users have concerning the RFID technology. Investigating several studies of interaction in ubiquitous computing, Marquardt et al. summed up critical issues from the view of the users, which are partially well-known in human computer interaction (HCI):

- Users failing "to realize the currently available privacy and security level of the system, often due to the lack of visibility of the system’s behaviour."[45]
- Users naïve mental models based on the line-of-sight principle[2].
- Users not being aware of the possibly large read distances of RFID tags
- Users not being aware of the constant availability of RFID tags
- Users perceiving RFID technology as a black box and having incorrect or no understanding of the inner workings of RFID technology. "In turn, this led to a limited understanding of possible security and privacy risks"[45]
- Users misleadingly assuming visual or auditive feedback by the RFID readers for every read activity, not anticipating that unauthorized readers will probably not be so kind to give any feedback.
- Users generally feeling powerless to protect themselves, due to a lack of control possibilities and knowledge.[25, 31, 47, 60]

In order to face these issues, Marquardt et al. refer to different privacy guidelines and framework solutions. According to Dourish et al.[17]"designers should allow decision making from within the application context itself"[45]. Further it is suggested to disable RFID tags by default[10]and to let the users decide when and to what extend information should be revealed[47]. Another one is the RAVEN framework by Bellotti and Sellen[5], which concentrates on feedback and control to find privacy issues and finding solutions for them. This actually goes back to the fundamental HMI approach that Donald Norman described with the seven stages of action in 1988[49]. Adapting these to privacy in ubiquitous computing Langenheinrich[37]devises six helpful principles: notice, choice & consent, proximity & locality, anonymity & pseudonomity, security, and access & recourse.

The evaluated RFID tag properties by Marquardt et al. form to the following. The core feature of RFID tags is an unique identifiability. Furthermore the ability to build them very small and embed them into other objects and materials cause an invisibility of tags. Due to that their status of activity is often not obvious, hence invisibility of use. In addition, RFID tags have a permanent availability, and in the past RFID tags were designed as stand-alone units without the intent to let users interact with or control the tags behaviour (autonomy).

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.10: Examplary Tags by Marquardt et al.[45]

Marquardt’s basic tag construction is made of a piece of cardboard, on which the circuit is installed with copper tape, the preprogrammed ICs and the remaining components like sensitive low voltage LEDs. All the tag prototypes communicate on a frequency of 13.56 MHz (see figure 2.10). The methods for improving the awareness of RFID tags include three types of tags. A visual feedback RFID tag simply provides an LED, which lights more bright, the more the tag enters the reader’s magnetic field. An audible feedback RFID tag and a tactile feedback RFID tag need an extra power supply via a small battery, making them semi-passive[25], because the current of the magnetic induction is not even sufficient enough to power the vibration motor or the piezo speaker alone. Similar to the visual tag, the audible tag starts buzzing and the tactile tag starts vibrating, when holding it into the magnetic field of the reader. On the other hand the methods for improving the control of RFID tags include nine tag types. The first one is a sliding switch RFID tag, which allows for toggling the tag on/off for a longer period of time, similar on how the so called "hold" functions in common mp3 player or SD cards work. For temporarily activating tags, Maquardt et al. propose three possibilities. A push button RFID tag and a pressure- sensitive button RFID tag, which only activate the tag while the button is pressed, and a touch-sensitive contacts RFID tag, which uses the conductivity of the human skin to activate the tag, when being hold in the palm of the hand. Further they present two tilt tags: a tilt-sensitive RFID tag that turns on and off depending on the orientation, and a flipping RFID tag, that switches between two different ICs enabling a multi ID tag. Furthermore, a light sensitive RFID tag, only turning on in lightish surroundings and two distant-dependent tags, a variable detection range RFID tag and a proximity-dependent disclosure RFID tag. The variable detection range is afforded by a slider that switches between different antenna lengths and therefore is directly controlling the detection range. The proximity-dependent disclosure tag consists of two separate tags on one cardboard, which only differ in antenna length. Like that the tag switches between IDs depending on the distance to the reader.

No particular study was conducted with the tags, but by publishing an instruction of the tags in the internet26 and holding a DIY workshop27 during the 2009 CHI conference, they gained qualitative responses and reactions about it. It still is a difficult aspect which degree of awareness and control will fit a certain application in the future. "The many activity notifications (e.g., an acoustic tone) could become annoying, and the required manual activation of the tag (e.g., repeatedly pressing a button) may become an irritant. The problem becomes even more apparent when considering the increasing number of RFID enabled cards people might possess in the near future. This could lead to situations where people prefer to disable the awareness and control mechanisms completely and instead fall back to an always-on RFID tag"[45]. Therefore, Marquardt et al. suggest to appropriately balance implicit28 and explicit29 methods depending on how sensitive the handled information is. By all means the last decision should always stay in the hand of the user, adapting the level of feedback and control to his individual needs.[45] The Work of Joseph Paradiso et al.

"Dr. Paradiso is a principal research scientist at the MIT Media Laboratory, where he leads the Responsive Environments Group and is the technology director for the Things That Think Consortium"[57]. Besides his 200 articles in the fields of sensor networks, energy harvesting and ubiquitous HCI etc.30, I have picked the two publications which are closely related to my research. The first one is a small self-powered pushbutton controller, the second one is about musical trinkets controlling music interface.

A Compact, Wireless, Self-Powered Pushbutton Controller Joseph Paradiso and Mark Feldmeier demonstrated that a wireless and self-powered pushbutton controller (see left picture in figure 2.11) is technically realizable. By pushing the button, an ID can be transmitted across the whole floor of a building. Their transmitter connected to a dipole antenna works at 418 MHz frequency. As an inspiration for the power supply they state the patent by A. Crisan, who invented a keyboard with little magnetic generators around each key, enhancing battery lifetime of laptops.[13]

The actual pushbutton is a piezoelectric element, taken out of an electrical lighter, which produces high voltages at low currents. To utilize this for normal electronic circuitry, they incorporated a step down transformer, which together with the right capacitor size creates an syntonic resonant circuit that powers a drop-out linear regulator. Unfortunately all these voltage stabilizations and regulations come at the expense of high inefficiency. But most important is that it’s working at all. The residual circuit just consists of a digital ID encoder (IC) and a RF transmitter.

A discovered disadvantage is that the hard strikes on the rigid piezoceramics cause the formation of microcracks, decreasing the pushbutton’s performance. Potential improvements are therefore proposed concerning the tuning of the piezoelectric element to the transformer together with the capacitor and the regulator, as well as improving the efficiency of each of these components. One last idea they suggest is that the piezoelectric generator could be better stimulated by using passive hydraulics at the element’s resonant frequency, like it is done by Antaki et al.[3].[51]

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.11: Pushbutton Controller by Paradiso and Feldmeier[51](left) and the Musical Trinkets by Paradiso et al.[53](right)

Tangible Music Interfaces using Passive Magnetic Tags Joseph Paradiso, Kai-yuh Hsiao and Ari Benbasat have build a music interface by using passive magnetic tangibles. The unique frequency values of RFID tagged toy figures were mapped to certain parameters of a music software. Different locations, distances and orientations of the so called Musical Trinkets produced different sounds (see right picture in figure 2.11). Therefore this is a realization of simultaneously read tags, which as well incorporate the distance of each tag. This is normally not really a trivial thing to do. Although the tags used by Paradiso et al. simply work on resonant (LC) or magnetostrictor basis, each one has an unique resonant frequency which is used for the identification process. This means they do not have any microcontroller working in a certain protocol, neither implementing an anti-collision software mechanism.

There are previous works and publications by Paradiso et al., which led to their final music interface design [50, 52, 55]. Big effort were spent on the development of the reader. In the beginning a pulse-induction ringdown reader was used, which "[sends] out a brief magnetic pulse at the resonant frequency of a tag and then [listens] for a ringdown response"[53]. Therefore all the tags needed different resonant frequencies. The reader dynamically tunes through the frequencies of the tags "by adding discrete capacitors through a triac-switched, binary-weighted ladder"[53]. The main problem of this design is the reading latency of 5-10 ms for each tag, which rapidly multiplies when using more tags at once. That is why they developed a second version, which is a swept-frequency tag reader.[56, 28, 57] Within this new and much more complex set up the dynamic loading of the tags is detected "by an inductive bridge, subsequently shaping and detecting the resultant signal through a bank of highpass filters."[53]The frequency can be sweeped between roughly 40-400 kHz.

The deployed series of 16 tagged objects include two toys31, which have three embedded tags, in order to enable orientation and proximity detection. Additionally a forth one is installed to recognize the position of the toy head32. And a few others can be attached to fingers, supporting a versatile controlling interaction experience. All objects were made out of nonferrous, non-conductive plastic or resin.

Even though I do not want to name all the exact mappings of all the objects to its specific sound parameter, I still want to mention a few. There is for example the proximity parameter controlling either the volume or the pitch depending on the objects. The pitch can also be controlled using the orientation of the cube. Furthermore there are few modifier tags for changing vibrato or adding other audio effects.

Despite the overall positive reactions, a downside of this music interface is the abstract mapping of the objects to the respective parameters, kepping people from starting right away. So do Paradiso et al. admit: "The mapping, although simple, isn’t entirely intu- itive, however, and this installation generally requires an attendant to get participants started"[53]. Last but not least they point out the possible benefits of a more uniform reader magnetic field. They already started experimenting with an arrangement of three perpendicular Helmholtz coils, accomplishing this way not only an improved identification, but also a precise tracking of objects within the interaction space.[53]

2.2.3 Sensor Enriched RFID

Similar to the musical trinkets by Paradiso et al.[54], the first wireless sensors worked by simply encoding the sensor data in an electromechanical manner. By manipulating the resonant (LC) systems characteristics, different response frequencies are produced. Thus, the highest sensor value correlates with the highest frequency within an unique bandwidth, for example. These sensor tags most often only worked because no microcontroller was needed to be powered together with the sensor. Among others, the work on chipless RFID sensors by R. Fletcher[20]is an example for these kind.

On the way to sensor networks with the RFID technology, Deng et al.[14]have simu- lated sensor-embedded RFID (SE-RFID) systems on electronic design automation (EDA) software, which helps during fast iterations in the early design stage. In more practical works in the field of sensor networks Mehmood et al.[46]lay out a grid of RFID tags on the floor, creating computer readable spatial information, a small remote car can be navigated upon. Furthermore Brunette et al.[8]have developed wearable, mobile and embedded wireless sensor nodes. Thereby contextual user information can be generated and an excellent holistic and coherent interweaving and interpretation of different sensor inputs is presented. Technically they rely on the Intel’s iMote (though not completely passive) and a custom hand-held reader. They finally suggest and demand platforms that "help blend sensor networks into ubiquitous computing"[8]and therefore also form the expression of "computations in blended sensor networks"[8]. Another example is the work by Potyrailo et al.[61]which experimented with passive 13.56 MHz RFID tags for diverse sensing applications. By combining "several measured parameters from the res- onant sensor antenna with multivariate data analysis, [the RFID sensors] deliver unique capability for multianalyte sensing and rejection of environmental interferences with a sin- gle sensor"[61]. They achieved to measure "toxic industrial chemicals (TICs) in air with the detection limit (DL) of 80 parts per billion"[61]and monitored the spoilage of milk. For the monitoring the supply chain of food like meat and fish, fruits, dairy and frozen products a few other sensor equipped RFID tags are already commercially produced, like e. g. the TempSense33 and different sensor tags by Phase IV34. In this context I also want to mention the possibility of an impact sensor, that could tell with one bit information, if a fragile product was damaged during transportation[80].

One of the most well known systems in the field of sensor networks with RFID certainly is Intel’s wireless identification and sensing platform (WISP) project. "A WISP consists of passive RFID tags augmented beyond the basic identification capability of ordinary RFID to support sensing and general-purpose computation"[70]It uses the UHF frequency band with the EPC Class 1 Gen 1 protocol and has a 16 bit microcontroller with an analog- to-digital converter. The sensor data is dynamically encoded into the ID, that the WISP is responding to an interrogating reader. That is why all of the 64 bit of the ID can be used to encode sensor data into a single read event. So far light, temperature, strain, tilt switches (also used as one-bit accelerometers) and rectified voltage sensors were examined and worked within read ranges of 4.5 m. Through the feedback of early experiments with the α -WISP, the idea of ID modulation35 was developed. This means the encoding of sensor information "by controlling the pattern of ID changes over time"[70]. For example, a tag that has a tilt sensor switching between two IDs, is able of providing one bit of information. In this case it makes no difference, whether the single bit is ’0’ and ’1’ or if it consist of two different IDs. The benefits of this kind of one bit encoding is that it universally works with all standard-compliant RFID protocols and needs no header overhead like parity CRC checks what so ever. "ID modulation allows us to create RFID sensor systems that work today, despite the fact that RFID sensing standards are not likely to be finalized for at least several years. Even when RFID sensing standards are in place, the ID modulation approach might still be used, since it allows system integrators to build solutions using the best (or least-expensive) available RFID tags and readers"[70]. Since the beginning of the α -WISP in 2005 a few WISP variations evolved that mainly concentrate on hardware improvements [59, 71], such as using a mini-harvester, powering the communication part and a main-harvester, powering the sensors[66]. Also one experiment used an extra large storage capacitor that can be charged by the RF energy and provides 24 hours power supply for the sensor node[9].

López et al.[43]incorporated wireless sensor nodes (WSN) with EPC Class 1 Gen 2 RFID tags. In a very promising framework they developed data models, sets of network protocols, tools to simulate and monitor the sensor network system and an IDE for their sensor platform as an Eclipse Plugin called ANTS EOS. Systems like these could be the foundation to also develop wireless and batteryless physical controllers with it.

Not only because of the commercial popularity of the EPC Class 1 standard, but also because of the scientific research that already relies on it, it stands to reason that the industry group GS1, that manages the EPCglobal standard, is working on protocols universally supporting all kinds of sensor technology and their different demands. In 2009 the German subsidiary of GS1 published a basic informations paper on EPC/RFID and sensor technology, describing a nearly finished EPC protocol standard for sensor technology. To my knowledge this standard has not found its way into officially released EPC protocol versions yet. So it will be interesting to see, what the future brings, but one should definitely keep one’s ears open concerning EPC’s possible sensor technology protocol.[15]

Another example is the work of Strömmer et al.[18], which have developed a multi- purpose platform, combining sensors with the NFC standard. This multi-purpose plat- form is called Smart NFC Interface and "is a light, matchbox-sized device with a micro- controller, rechargeable Li-Ion battery with charging electronics, data logging memory, RS232 serial port and other wired communications, as well as NFC, IrDA and Bluetooth wireless communications"[18]. The smooth integration of sensors comes at the expense of the downside that, in contrary to the earlier presented WISP, it is not completely passive and is in need of a battery for power supply. The platform device is split into a commu- nication board for the NFC communication and a basic board responsible for providing an interface to different sensor types. These two boards themselves are connected via one port called board-to-board connector. Problems have been the handling of the micocontroller’s different power modes and to this effect also getting their wakeup interrupts to work.

One of the technically most sophisticated works is written by Cho et al.[11]. The team was able to build a UHF RFID tag with integrated sensors for environmental monitoring, that dissipates only 5.14 μ W during active state36. The main trick enabling such a low power consumption is a "fully integrated clock generator independent [of the] external RF signal"[11]. In the UHF band, frequencies between 869-920 MHz are predominant, but letting a microcontroller work at these frequencies means higher power consumptions compared to common microcontroller frequencies of i.e. 4-20 MHz, which are still sufficient enough to provide acceptably computational speed.

The reason why the research in all these sensor network possibilities is substantial for this thesis, is because of the inconceivable great potential for future HCI. Strömmer et al.[18]put it on point: "In short-range radio based wireless sensor networks, there is not necessarily a human user at all, but the sensors may serve as a part of an automation system, for example. In contrast, a human user plays a central role in typical NFC-based implementations, a mobile handset such as a mobile phone being the typical tool for the interaction." With smart phones being the future service access point to contextual in- formation of the sensor networks, models for implicit HCI and proactive computing (cf. Schmidt[67]) seem to get a promising technology to become real life implemented. In this sense the work in sensor networks provides a superior model, that also applies to the topic of wireless and batteryless physical user interfaces. Instead of sensors continuously mea- suring e. g. environmental conditions, control elements measure human user inputs, which in perspective of a back-end system only means a certain category of input information. The technological low-power infrastructures and inventions will likewise suit for sensors and control elements.

Interactive RFID Prototypes

An extensive work of this thesis, was the building of several interactive prototypes, that all rely on different RFID technologies. This also means, accomplishing multiple entry tun- nels, in order to achieve a state of reliable operation. The goal of building these prototypes is the evaluation of their technical benefits and limitations. The first set of prototypes relies on unmodified RFID chips. The second set of prototypes relies on programmable chips and hence extends classical RFID chips with additional computational processing. To the end I investigate the properties of RFID antennas for both prototypes and conclude in selecting the appropriate technology for realizing a first Tesla User Interface.

3.1 Paper RFID Tags

In the first series of prototypes, I rebuilt some of Marquardt’s tags (see section The Work of Nicolai Marquardt et al.), to gain first qualitative impressions on the functionality and the usability. Sticking to the dimensions and the basic schemes it is really easy to build working paper RFID tags right away. Before starting with the tags, one has to get all the necessary components and should have a working RFID reader, for being able to test the correct operation of the built tags. This complete process of building and testing is documented in the following section.

3.1.1 Components

This section explains the single components of the paper RFID tags and an according 13.56 MHz RFID reader and their assembly, in order to read IDs with an standard computer.


1 http://www.bdi.eu/download_content/EnergieUndRohstoffe/Faxtsheet.pdf, retrieved 17.03.2013

2 see section 2.2.3 for details.

1 http://www.technologyreview.com.br/printer_friendly_article.aspx?id=17182, retrieved 07.03.2013

2 http://en.wikipedia.org/wiki/Moore%27s_law, retrieved 07.03.2013

3 http://www.emmicroelectronic.com/webfiles/news/EDNRFunlocked.htm, retrieved 07.03.2013

4 http://www.transcore.com/products/rfid/rfid_readers.shtml, retrieved 07.03.2013

5 Here e. g. in nice compact set up http://www.cooking-hacks.com/index.php/shop/arduino/

arduino-rfid-pack.html, retrieved 07.03.2013

6 KTS Systeme, RF RFID read/write USB sticks https://www.buerklin.com/default.asp?event= ShowArtikel%2813M2945%29, retrieved 07.03.2013

7 http://www.freetronics.com/blogs/news/2280582-rfid-access-control-using-ethernet-poe# .UQqiqGdbUw8, retrieved 07.03.2013

8 http://www.trossenrobotics.com/p/REDBEE-Wireless-rfid-reader.aspx

9 among others used e. g. by the EPCglobal Class 1 Generation 2 standard

10 also called Manchester code definition by G.E. Thomas or simply Manchester II

11 published under the Creative Commons Attribution - Noncommercial - Share-alike 3.0 Unported Li- cense.

12 The business point of view derives from being geared to supply chain management and logistics as the main field of application, not taking ubiquitous computing and new user interactions into account

13 Due to format reasons the complete taxonomy can not be displayed in full size in this written thesis. For a closer look onto the big picture that RFID is today, please visit my according thesis blog entry at http://teslaui.wordpress.com/2013/01/25/2013-rfid-taxonomy/, retrieved 07.03.2013

14 complete list of all ISO/IEC standards that somehow affect RFID: ISO 11784, ISO 10374, ISO 11785, ISO 14223, ISO/IEC 14443, ISO/IEC 15961, ISO/IEC 15962, ISO/IEC 15693, ISO/IEC 18000, ISO/IEC 18001, ISO/IEC TR 18046, ISO/IEC TR 18047, ISO 18185, ISO/IEC 19762, ISO 23389 , ISO/IEC 24710, ISO/IEC 24729, ISO/IEC 24730, ISO/IEC 15434, ISO/IEC 15459, ISO/IEC 24752, ISO/IEC 24753, ISO/IEC 24769, ISO/IEC 24770

15 Class 0 (900 MHz), Class 1 (13.56 MHz, 860 MHz - 960 MHz), Class 1 (Gen2) (860 MHz - 960 MHz), Class 2, Class 3, Class 4, Class 5 (http://www.rfidjournal.com/articles/view?1335, re- trieved 07.03.2013)

16 cooperation of universities and researchers placed at the MIT and former called Auto-ID Center

17 The Zigbee Alliance manages the ZigBee protocol working at 868 MHz and 2,4 GHz frequencies

18 In the meantime officially replaced by EM4200

19 In the meantime officially replaced by ATA5577

20 HID Global manages i.a. the iClass protocol (13.56 MHz), SmartID protocol, HID Prox protocol (125 kHz), the Indala Proximity protocol and some more: http://www.hidglobal.com/sites/hidglobal. com/files/hid-idt-products-chart-en.pdf, retrieved 07.03.2013

21 Classic 1K, Classic 4K, Ultralight C, Ultralight, Plus X, Plus S, DESFire, DESFire EV1, SmartMX

22 e. g. FasTrak for the San Francisco Bay Bridge

23 Oyster Card in London

24 i.e. Google ’s Nexus 4 by LG released in Germany in November 2012

25 The semi-passive tags are not that interesting for the work of this thesis, because exactly the self- sustaining nature of completely passive RFID tags is one of the goal key advantages, which lets them be the tool to realize the ubiquitary dream of the "internet of things".

26 http://www.instructables.com/id/RFID-Reader-Detector-and-Tilt-Sensitive-RFID-Tag/, re- trieved 07.03.2013

27 http://grouplab.cpsc.ucalgary.ca/grouplab/uploads/Publications/Publications/ 2009-RFIDReader.CHIWorkshop.pdf, retrieved 07.03.2013

28 implicit could just be a consistent way of using RFID tags

29 e. g. could be inconsistencies (like with the formatting c command and ok not being the default button) gain users consciousness for a severe not returnable action

30 http://www.media.mit.edu/people/joep, retrieved 07.03.2013

31 cube and eyeball

32 a PEZ dispenser

33 http://www.alvinsystems.com/resources/pdf/MobileRFIDSensor.pdf, retrieved 07.03.2013

34 http://www.phaseivengr.com/p4main/Technology/SensTAG%E2%84%A2WirelessRFIDSensors.aspx, retrieved 07.03.2013

35 also called on-off keying (OOK)

36 As a comparison the Smart NFC Interface just presented before needs approximately 30 μ W18 and one of the latest designs of the WISP needs approximately 10.8 μ W (600 μ A for one EPC read query at 1.8 V 600 μ A · 1.8 V = 10.8 μ W)66

Excerpt out of 120 pages


Tesla User Interfaces
Wireless and Batteryless Tangible Interaction
LMU Munich  (Institut für Informatik)
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
6609 KB
This thesis documents the research and development of new interaction concepts for passive sensor RFID technologies. Therefore it first of all consistently structures and categorizes the whole technological field of RFID. In the end it opens up possibilities for interaction concepts with the passive sensor RFID technology. In the addendum a vast list of related scenarios provides a fast introduction into the possibilities of this technology.
RFID, HCI, Ubiquitous Computing, Low-Power, Antenna Tunning, Microcontrollers, Sensors, User Interfaces, Taxonomy
Quote paper
Dario Soller (Author), 2013, Tesla User Interfaces, Munich, GRIN Verlag, https://www.grin.com/document/232570


  • No comments yet.
Read the ebook
Title: Tesla User Interfaces

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free