Place and product-based collaborative filtering

In March 2006 fourth year interaction design students at AHO conducted intense one-week investigations into Near Field Communication in a project called Touchable services. See more student projects.

Jon Olav Eikenes, Guilia Schneider, Bjørn Erik Haugen and Marie Wennesland created a high-level concept that proposed the idea that once we start to use our phones to pay for things, couldn’t we also start to use collaborative filtering of places and purchases?

touchable_service05.jpg

As ‘wallet phones’ become commonplace across Asia, we need to think about how these applications will be created and marketed here in Europe. What ‘added value’ will these phones offer over cash or credit cards? In this scenario, the ‘wallet phone’ not only let the user know how much they have paid, but it will recommend and offer social commentary on their choice of both products and places.

touchable_service06.jpg

Building an ‘Amazon’ for places and products is interesting because it builds upon everyday transactions and behaviour. The data that users and communities could pull out of these transactions is potentially very rich and useful. It also raises privacy concerns, but with the focus on users or communities owning and managing their own data on their personal devices, this project advocates for user-agency in such a system.

More details and images at Jon Olav’s & Giulia’s weblogs.

Posted in Interaction design, Mobile, Product design, Projects, Research, Retail, Service design, Student projects | Tagged , , , , , , | 2 Comments

Touchable services: local interactions

touchable_service01.jpg

In March 2006 Fourth year interaction design students at AHO conducted intense one-week investigations into Near Field Communication in a project called Touchable services. See more student projects.

Einar Sneve Martinussen, André Borgen, Paolo Dell’elce and Henrik Marstrander looked at how to increase the cohesion of local communities. As a starting point they studied a local record shop and looked at the intricate social and economic relationships around it.

touchable_service02.jpg

They discovered an existing layer of printed, handwritten, and scrawled information in and around the record shop itself. This was a mixture of content, relying very much on local knowledge and social connections including recent arrivals, staff picks, recommendations and playlists, all of which offered rich opportunities for interaction design. They also discovered intruiging patterns and behaviours, including eccentric opening times that inspired them to look at what happened around the shop during closed periods.

touchable_service03.jpg

They prototyped a simple application that offered the ability to hear and download music, playlists, new releases, etc. by touching the phone to the shop window, and conceptualised how this might be applied in other areas of the shop.

touchable_service04.jpg

See more at Henrik’s weblog with their presentation of technologies, wider ideas and research.

This project was very interesting in the amount of ideas that emerged from a single context. It seems that when you begin to research a specific situation, the applications and services that emerge from it are numerous. The students here conducted a kind of situation-based brainstorming that was very rich. Situation and context provide very useful limits for for idea-generation. It also showed that NFC has numerous opportunities in niche communities, in contrast to typical location based services that offer generalised applications to tourists, etc.

Posted in Interaction design, Mobile, Projects, Research, Retail, Service design, Student projects | Tagged , , , , | Leave a comment

RFID hello world

Phidgets are go

In what must be one of the quickest ‘Hello world’ experiences, the Phidgets RFID reader was up and running on OS X in about 2 minutes after unpackaging. Phidgets must be congratulated for making such accessible and immediate physical computing hardware and software.

Detected tagMetal + RFIDs

A simple Cocoa application reading RFID serial numbers, and playing with RFID tags obscured by metal objects. Having multiple, numerous readers around gives us the ability to quickly play around with objects, spaces and tags: understanding the range, directionality and responsiveness of RFID.

Posted in Research | Leave a comment

Phidget RFID

Phidget RFID reader

Ten Phidget RFID boards arrived today. The Phidgets are an extremely cheap prototyping platform for simple RFID. They are low power, with a maximum read range of about 8 centimetres, so perfect for quick interaction prototyping of near-field applications. We’re planning to build some table based applications, and to use them in student projects. The only downside is the lack of anticollision: multiple tags can’t be in the vicinity of the reader at once.

Phidget RFID readerPhidget RFID reader

The boards plug in via USB to a host computer (can be used with USB hubs to create massive arrays of readers). They have drivers for Windows, Linux and OS X, plus libraries and sample applications for Visual Basic, C, C++, Flash, .Net, Java etc.

With the readers came a huge variety of low-frequency tags, many of which are interesting from a design perspective. The CD/DVD tag, the tough polycarbonate buttons, and the very pedagogic clear plastic tags.

RFIDs

[tags]RFID, Phidgets, prototyping, electronics, physical computing [/tags]

Posted in Technicalities | 2 Comments

Phantom geometry and tissue-simulant liquids

Indexsar's products for testing emf

I’m investigating the visualisation of electromagnetic fields, part of our exploratory process to look at the materiality of RFID. What are we talking about when we say ‘touch-based interaction’ or ‘near field’ for instance? This investigation threw up an interesting company: Indexsar specialises in:

“Turnkey test systems for the measurement of SAR (Specific Absorption Rate) and the Over The Air (OTE) testing of wireless devices. Our product range includes E-field probes for use in both air and in tissue-simulant liquids, equipment for accurate testing of the dielectric properties of phantom liquids and a system to give a 3-Dimensional presentation of mobile handset radiation and sensitivity. We can supply suitable RF amplifiers and directional couplers for wireless product testing and can offer suitable phantoms (heads, hands and liquids) for testing radiated emissions.”

Their test setup includes 6-axis industrial robots to move sensors around models of human anatomy (or “phantom geometry” in official testing language). Their setups include geometry for the head, and the right hand (no left hands yet). In the test rig below a liquid model moves while the sensors are stationary. It looks like it has been hacked together from meccano and a Wacom pad.

Indexsar's test setup for testing emf

These are some visualisations of bodily electromagnetic fields from Flomerics MicroStripes and Hugo:

Visualisations of EMF in anatomical models

Microstripes software is a 3D electromagnetic simulation & synthesis tool:

“MicroStripes is widely used to design antenna and microwave structures and assess their installed performance, to optimize RFID systems, to analyze radar cross-section (RCS), EMI/EMP and lightning effects on vehicles, ships and aircraft, and to predict absorption of EM fields in human tissue.”

Another visualisation:

Visualisation using Flometrics MicroStripes.

Leafing through the manual for my Nokia E60 I noticed that it includes guidelines on holding the phone, in order not to degrade the performance, and thus battery life of the antennae.

Nokia E60 manual

As we move towards multiradio devices, this analysis is going to become more important. I wonder what the fields look like around the RFID/NFC phones like the 3220...

Posted in Mobile, Research, Technicalities | Tagged , , , , , | 4 Comments

RFID in Seoul: High-end smartcards

The contactless ticketing and payment system of choice in Seoul is called T-Money. Seoul was the first city to use Mifare standard smartcards in 1996. Although retail payment doesn’t seem to have taken off as much as in Japan or Hong Kong, T-Money is fairly ubiquitous and can be used on all public transport, a few vending machines and selected retailers.

There are two elements to this system in Seoul that are interesting. The first is the availability of customised cards and accessories from market stalls and groceries (the subject of a future article). The second is the availability of high-end products containing the T-Money smartcards. These are designed to compete with other products in categories such as consumer electronics, fashion, toys, music and computing.

Here are three examples of this type of product.

RFID Hippo

RFID hippo (Tmoney)

This ‘Fancy T-Money’ hippo costs 24,800 Won (about €20) available from the online shop. It is a smartcard with a USB interface, which means that you can plug it into your PC and fill it up with money from an online bank account (after installing the right drivers). Inside the hippo you can feel a standard 56×28 mm RFID card that is connected to the mouth via a cable.

The USB plug in this case is thoughtfully placed in the hippo’s mouth, which means it looks like the hippo is chomping away at the computer when plugged in (not as grotesque as the USB bear or as weird as the USB twig).

RFID hippo (Tmoney)Fancy Tmoney

Given the Korean love of characters, I’m wondering if the hippo is becoming a symbol or mascot for T-Money, in the same way as the Suica Penguin has become synonymous with mobile payment in Japan. Or will this be the start of a collectible range of characters? Some T-Money is bought with discounts built in (for kids or elderly for instance), it would make sense that this is for children, but I’m not certain that this is the case. Let me know if you know more. More photos at Flickr.

T-Pop

T-Pop USB reader/writer (Tmoney)

This is a tiny RFID reader/writer that is bought alongside a standard T-Money card for 9,900 Won (about €8). Plugging the T-Pop into the USB port and placing the card on top allows it to be topped up. The T-Pop is designed with a clip so that it can be carried around, but the clip also acts as a holder when the card is being charged. This might have been a nice compromise on the drawing board, but in practice it seems clunky: the action of charging a card could be embodied in the physical form, rather than as an afterthought. This is interesting as the first consumer RFID reader/writer that I have found.

T-Pop USB reader/writer (Tmoney)T-Pop USB reader/writer (Tmoney)

What does putting the RFID reader/writer in the hands of users do to the service in terms of hackability? One would think that both the near-field and USB transfer of cash could be analysed in detail and reverse engineered in the comfort of one’s own home (it’s very different from trying to analyse transactions in the wild).

+Plus

RFID memory stick (Tmoney)

This last example is a simple memory stick that also acts as a T-Money card, a 256 Mb version is available for 39,800 Won (about €32). As the designers began to explore the recharging of cards via USB it must have seemed obvious to integrate other USB devices. Thoughtful convergence means that topping up money on this kind of T-Money can be done at the same time as transferring or using files.

RFID memory stick (Tmoney)RFID memory stick (Tmoney)

There are other convergent examples, some of which I will cover in the future. The T-Money mp3 player, also available from the online shop, was beyond my budget.

New service touchpoints

Overall the re-packaging of contactless cards is an interesting mix of service, interaction and product design. The way in which the service touchpoints of the Seoul transport system are fanning out to include the home PC is interesting. By providing software and hardware for charging smart cards at home, T-Money have invested a lot in technical and service development. What kind of studies went into making these products: who said that they needed to top up their cards at home? Would the same service work in other places, with the Oyster card for instance?

Convergence

It’s difficult to know the relative success of these convergent devices (MP3 players, memory sticks, even mobile phones). My contention is that the usability benefits of these converged things do not outweigh the drawbacks of using a compromised device, or having options in style, interface, or any of the other things that people like to have choice in. When T-Money cards can be attached to phones, cameras or any other device as straps, embedded within wallets or handbags, or worn in sleeve pockets, the arguments for convergence seem quite weak. Products like the hippo show us that many functions will remain discrete.

Materials & product design

Products like the hippo are just scratching the surface of possibility, particularly around the use of materials and the kinds of emotional attachments we have to everyday objects. The underlying technology (RFID) allows easy re-packaging in almost unlimited forms and materials (as long as it’s not metal or liquid). I’d like to see explorations of high-end materials like leather, glass, stone and ceramics, to look at reliability, trust and to think more about the ways in which these functional things could be inflected by fashion, design and everyday behaviour.

What would the ‘ideal’ payment device look like? Does the fact that these things are designed for public transport place them within a certain category of objects that are mundane and everyday, or even within a certain social status?

Posted in Interaction design, Product design, Retail, Service design, Ticketing | Tagged , , , , , | 1 Comment

Nordichi workshop papers

Update The papers are available to download as PDF.

Nordichi paper images

A selection of images from submitted papers, click for larger image.

The diverse range of topics and the varied backgrounds of the applicants for our Nordichi workshop in October is promising for a topic that spans architecture, hci, computer science, interaction and industrial design. We are really looking forward to seeing everyone in Oslo.

These are the 15 accepted papers:

Bootstrapping the Internet of Things
Claus Dahl, Imity

CybStickers - Simple Shared Ubiquitous Annotations for All
Odd-Wiking Rahlff, Sintef. (See also Cybstickers)

Designing Expressive Near Field Interactions
Johan Sandsjö, Hidden Interaction

Designing social affordances for material objects
Ulla-Maaria Mutanen & Matt Biddulph, University of Helsinki (See also Thinglink)

Everyday Intelligence
Gill Wildman, Plot (See the Everyday Intelligence film)

*The Kinetic User Interface *
Vincenzo Pallotta, Béat Hirsbrunner, Pervasive and Artificial Intelligence Research Group, University of Fribourg, Switzerland

Mobile Prosumer
Florian Resatsch, Stephan Karpischek & Daniel Michelis, IEB, University of Arts Berlin

PERvasive serviCe Interaction
John Hamard, DoCoMo Euro-Labs

RFID work
Morten Borup Harning RFIDsec

Responsible design of connected objects
Alexandra Deschamps-Sonsino

Security Aspects in Design of Touch-based Applications
Janne Jalkanen, Nokia

Social Construction Kits for Kids, Digital Infrastructures for Pervasive Play
Martin Brynskov, Interactive Spaces, University of Aarhus

Unconscious Kitchen, Everyday domestic behaviors & Blackbox, Public Energy Awareness
Joseph Yang

Use of Near Field Communication in emergency Rescue situations
Gunnar Kramp, Aarhus School of Architecture

Waschsalon mobile service
Chris Woebken

You can only touch what is there
Chris Heathcote, Nokia

We also had expressions of interest from Willem Velthoven with the Symbolic Table and Rob van Kranenburg. We’ll keep in touch and hope you can attend.

Posted in Events, Workshops | Tagged , , , , | 4 Comments

Developing NFC applications

Another NFC shell

Judging by the number of emails we have received, there is great interest in software and hardware development of NFC applications. The main stumbling block at the moment seems to be working out what development platform to use, and knowing where to get hold of development kits for the various handsets and servers needed to run an NFC application.

Bearing in mind that we haven’t used any of these things extensively yet, we are putting everything we know about getting hold of development resources here. Perhaps we can collectively use this post as a way of pooling information on the topic. Thanks to Peter Ljungstrand for initial conversations and valuable information on these things.

NFC phones

At the moment, the phones we can use for NFC development are the 5140 Field Force Solution and the 3220 NFC Shell. Both of these are rather quirky, the the 5140 is slow, and the NFC shell is somewhat unreliable. Nokia has admitted that these are trial releases for developers, and will never be widely released to the public.

Nokia Field Force phone

Both of these kits are available from Top Tunniste, and we’ve seen them pop up at other online stores too, so search around.

Various rumours have been circulating about upcoming Nokia handsets: perhaps a flip phone with a fully integrated NFC chip, not just an interchangeable cover. We’ll have to wait and see but lets hope they have a robust and consistent approach to the next release, coupled with simultaneous rollouts of useful public services.

Samsung has apparently developed an NFC enabled SGH-X700. There is some discussion about it over at the Sun developer network.

In Asia, particularly Japan, many handsets are using the Felica standard that is supposedly compatible with NFC. It would be very interesting to hear any Asian experience with these phones or the Felica SDK.

NFC tags

The Nokia 3220 only comes with four tags, so any developer will need to get hold of more. Getting hold of tags is easy. Toptunniste for instance sells a few kinds of tags for about a euro each. But there are many other options: NFC is based upon high frequency ISO 14443 Mifare standard, so re-writeable Mifare Standard 1K/4K or Mifare Ultralight tags from any supplier should work. And these are available in many form factors, from cards, key fobs, stickers, to laundry tags and ‘mount on metal’ tags. Some other suppliers are Mannings, Promobox, Cardxx, SAG and Korea RFID.

NFC standards and protocols

The NFC forum is slowly releasing the specifications for NFC devices, and this is the first place to look if you want to find out if the NFC standard is right for your application. There is enough detail there to work on high-level descriptions of NFC applications and services.

‘Official’ development kits

At the moment, because the Near Field Communication specifications haven’t been normalised, writing an application for NFC requires writing for one brand of phone, using proprietary development kits.

Nokia Field Force Architecture
The Nokia Field Force Solution Architecture, available here.

The Nokia development kit for NFC is currently available from a number of suppliers. Top Tunniste in Finland sells a 5 user license for €2000, they also sell phones and compatible tags if you can’t find them locally. I’ve heard that the kit is available for around €800 in Germany and £500 in the UK so it may be worth phoning around to ask local suppliers.

In the license agreement for the Nokia SDK you agree to use their proprietary server for networked interactions. For this you need to buy, or subscribe to the Nokia Local Interactions Server (LI Server). The LI server is described as such:

Nokia Local Interactions Server (LI Server) is a cost-efficient, real-time Web service that simplifies data capture, reporting, management, and communication with mobile workforce. Quickly integrated into your company’s back-end systems, your field force personnel can use it to provide up-to-date information to the back-end office and receive instant feedback in their mobile phones. It is also an easy, but secure method to distribute and update company-specific phone Local Interactions (LI) Client software to end-users’ mobile phones.

This is €400 from Top Tunniste. Getting around the LI server is supposed to be a little clumsy, but it’s possible (if not legal) to use your own server applications. At the simplest end of the spectrum, my students have used URI stored on tags as simple pointers to web applications that are invoked when a page is loaded, for instance.

If you want to discover how other people are getting on with this stuff, then try searching at Forum Nokia for RFID or NFC, which reveals threads like this one.

Other development kits

There are other hardware and software NFC development kits available from places like Sirit, Inside contactless, Skyetek, Innovision and Wireless Dynamics. Given that the NFC forum specifications are still nascent, we don’t yet trust the interoperability between different manufacturer’s devices. So we haven’t used any of these so far, and would be very interested in your experiences.

Contactless communication API

JSR 257 the “Contactless Communication API” has been proposed by Nokia and is currently being defined by the Java Mobile community. It is supported by the Nokia 5140i FFS (Source) but we’re not sure about the Samsung phones. When rolled out and standardised, it should be a good common development space for all kinds of applications that require physical links or applications. One of the great possibilities is that we can prototype and launch using 2D barcode tags, and easily transition to NFC interactions as the handsets become widely available. We can also design more specialist applications that use Bluetooth or even IR in the same environment.

Simon Woodside, the developer of Semacode, writes about it here:

JSR 257: Contactless Communication API is defining both RF and visual tag interface for Java ME. This will be THE way to interact with visual tags and RFIDs, smart cards, etc. And since our SDK implements the public draft interface, you can future proof yourself.

There is an overview of this and JSR 256, the “Mobile Sensor APIhere, and an in-depth look at the standard here. It’s currently in the “Proposed Final Draft” stage so it looks like it’s pretty solidified.

More information

We are really interested in hearing about your experiences with development, it would be great to use this post as a place to collect information.

Posted in Technicalities | Tagged , , , , , , , , , , | 8 Comments