UWB Touchless Express Transit and Apple Pay for iOS 15?

A recent sudden surge of hits from Hong Kong accessing my December 2019 UWB Touchless Mobile FeliCa post seemed odd. I dug around and it appears that Hong Kong MTR, like JR East, is making noises about incorporating UWB technology in next generation transit gates.

iOS 14.5 added a new PassKit call for Bluetooth and the U1 chip integration since iPhone 11 and Apple Watch 6, coupled with global FeliCa support certainly puts Apple ahead of the game. I have no idea what WWDC21 will deliver but more UWB integration is a given.

Apple only mentioned UWB Touchless at WWDC20 in connection with digital car key without showing anything because the Car Connectivity Consortium Digital Key 3.0 spec was a work in progress. Now that the spec is in-place with BMW said to deliver car models incorporating UWB Touchless this year, will Apple show it in action? I think it’s highly likely, but since Car Key is a ‘Wallet Card’, and Wallet app Express Cards come is 3 types: Transit, Student ID, and Car Key, the more interesting question is…will Apple also show Touchless Transit and Student ID Express Cards? And what about Apple Pay?

People think Touchless is a completely new thing for ‘keep smartphone in pocket’ transactions, and they worry about security. You can’t blame them because marketers are selling the in-pocket payment experience. However, Touchless is simply long distance NFC without NFC. All UWB Touchless does is describe the frequency to use Bluetooth instead of NFC. The background stuff, secure element and so on, is exactly the same. This means user interaction is the same. For walking through transit gates and security doors, or unlocking your car, the convenience of Touchless is easy to understand: no more NFC tapping, just keep moving.

What about Express Card payments? The current Apple Pay Suica payment checkout experience: the user taps Suica on a touchscreen, or tells the clerk “Suica” then holds the device to the reader. The user has to give consent before the transaction is activated by checkout staff or the self checkout reader. For Apple Pay EMV transactions users have the extra step of confirming a transaction by Face ID/Touch ID to complete it.

Realistically however, in what situations does Touchless make store checkout more convenient and faster? Drive thru certainly, supermarkets…maybe, but most stores will probably not want to invest in Touchless without a good reason when the NFC readers they already have installed get the job done. There is one more interesting role that Apple has planned for UWB however, one that promises to improve the entire Apple Pay and Wallet experience: communicating with the reader before transaction to select the right Wallet card for the job, at a distance, for a truly smart Wallet app. With national ID cards, passports and more coming to Wallet at some point, UWB could be the Wallet reboot we really need.

And then there is EMVCo. The problems with UWB Touchless for EMVCo are that: (1) Touchless only works with devices with batteries, á la AirTag, and doesn’t work with the current plastic card model, (2) UWB + Bluetooth level the digital playing field with FeliCa and MIFARE, no more ‘real’ vs ‘who cares’ NFC hardware flavors to split hairs over. The plastic card NFC limitation is probably a bitter pill for everybody but especially for EMVCo members and issuers as plastic card issue is big business, and many customers are more comfortable with plastic cards. For those reasons I think EMVCo will be the last to support UWB Touchless, if they do at all. On the plus side Touchless does give digital wallet platforms an edge to create smart aware wallets, digital does NFC and Touchless, plastic only does NFC. We’ll find out about Apple’s UWB Touchless roadmap at WWDC21.

Octopus 2.0

The Apple Pay Octopus launch in June 2020 marked the end of an era of Octopus as the exclusive Hong Kong MTR home grown transit platform, and the start of MTR integrating into China mainland transit fare standards. In August 2020 Octopus Cards Limited announced they would join China T-Union. My take about it and the eventual migration of Octopus from FeliCa to PBOC 2.0, struck a raw nerve and did not go down well with some Hong Kong folk:

Can someone tell the ill-informed, self-centred, attention-seeking blogger to stop spreading fake rumours about octopus ditching FeliCa? Not in this lifetime…The self-proclaimed expert blogger’s been wrong on so many levels I’m amazed people still follow him like religion and never question his fantasy stories. Utterly annoyed by him dropping quotes from people out of context and use them to his benefits.

In April 2021 new OCL CEO Angus Lee Chun-ming said in a South China Morning Post interview that OCL had applied for China T-Union membership as planned, and will launch a dual mode Octopus card for mainland transit use:

“We have applied to join the China T-Union, the nationwide one-card payment system led by the Ministry of Transport. That will enable Octopus physical-card holders to pay for public transport fares in mainland China,”…

The service can be upgraded to digital Octopus cards in the phase two development. “The card will be denominated in Hong Kong dollars. Octopus will arrange the currency settlement with the mainland partner,” said Lee.

A one-card nationwide payment system eh? Sounds like an plug for China T-Union instead of an Octopus presser. Phase 1 is a physical dual mode Octopus card that appears to be 2 separate chips (PBOC and FeliCa) in one card with a common HKD ePurse. This is novel as Greater Bay Area dual mode cards up to now used separate ePurses for each currency. It’s also complicated because mainland transit operators have to do the currency conversion. A digital wallet version is phase 2. The elimination of FeliCa on the Hong Kong side will be the final phase, though that depends on the Ministry of Transport removing the current PBOC restriction that limits it to transit use and T-Union branding issue, or Octopus coming up with something else. We shall see.

On the mobile side Hong Kong iPhone users already have a dual mode Wallet option to add China T-Union cards if they have a China UnionPay credit or debit card. It’s not dual mode on one card and there is an Express Transit issue when turning on a China T-Union card turns off Express Transit for Octopus, but it works.

Dual mode transit cards on Apple Pay don’t exist yet but they are technically possible. Apple Pay already uses dual mode NFC switching for Japanese issue payment cards, FeliCa for contactless use in Japan, EMV for contactless use abroad. Another option might be the multiple secure element domain/multiple NFC protocol support of Mobile FeliCa 4.1 outlined by FeliCa Dude for dual mode transactions using just Mobile FeliCa with NFC-A/NFC-F.

On the transit gate side it will be interesting to see what design MTR uses for multiple protocol open-loop. NFC requires the reader side to specify the NFC protocol used for the transaction. This is a not a problem at store checkout, but how does the user specify the transaction protocol on transit gates? Answer: by tapping different readers. Perhaps the new MTR gates will host a NFC-A reader (EMV and PBOC), a NFC-F reader (FeliCa) in addition to the already separate QR reader? And if those Touchless UWB rumors are true, UWB and Bluetooth could be joining the MTR next generation gate party. One thing for sure, transitions are messy, and expensive.

Real world iOS 14.5 Face ID Unlock with Apple Watch performance

Now that iOS 14.5 is nearing official release, it’s time to check in on how far Unlock with Apple Watch for Face ID with face masks has improved over the beta testing cycle. The good news: Unlock with Apple Watch performance has improved from iOS 14.5 beta 1, the bad news: not so much. It still feels like a stopgap, it fails too often and Apple Music playback still hiccups with every unlock attempt.

I think performance will vary, a lot, depending on the user, the mask and the environment. For some, perhaps the majority, it will be enough. I find it fails me too often on the daily commute and in stores, usually at the very moment I need to launch dPOINT or dPay apps at checkout. I also get the feeling that Apple Watch battery life takes a hit too, but take it with a grain of salt along with my impressions. If it works for you that’s great, but the Unlock with Apple Watch end user experience will be all over the place.

End of the line for Suica and the native Japan Transit IC smartcard standard?

There is a consistent theme among some Japanese tech journalists: the native Japan Transit IC smartcard system is obsolete and destined for that fabled junk heap, the Galapagos island of over-engineered irrelevant Japanese technology. The arguments always boil down to cited higher costs of maintaining the ‘over-spec’ proprietary FeliCa based inflexible transit IC architecture in face of ‘flexible, lower cost’ proprietary EMV contactless bank payment tap cards and smartphone digital wallets used for open loop transit. Is Suica really ‘over-spec’ or is it clever stealth marketing sponsorship from EMVCo members and the bank industry disguised as journalism? Logically the same argument applies to proprietary MIFARE smartcard transit systems as well but is never mentioned, presumably because it was invented in Europe instead of Japan.

Despite all the digital ink on the subject I have yet to see a single article where said costs are actually shown and compared. Smartcard deposit fees are a standard way to offset plastic issue costs and Japanese transit companies like to earn interest off the float of card deposits and unused stored value. But this is never discussed nor the fact that digital wallet issue is free of hardware costs.

Bank payment cards and smartcards have very different business models. EMVCo members and their card issuers can hide associated hardware and licensing costs in bank transaction fees that NXP, FeliCa Networks and other smartcard technology solution providers cannot. Without hard numbers we can only take journalist claims at face value, that transit smartcards are not smart at all, but expensive obstacles to lower cost open loop centralization nirvana.

I don’t buy the ‘one solution fits all’ argument and neither should you. One constant issue in our internet era is that too much centralization is not only a technology monoculture security risk, cloud services fail, and cloud centralization is abused to limit human rights. As speech is censored on SNS platforms and online profiling is used to limit freedom of travel with politically biased no fly lists, it is inevitable that face recognition transit gates will be used to track people and implement ‘no ride’ or ‘limited ride’ policies. These are issues that people must be aware of in the relentless rush towards online centralization of transit payments and services.

Nevertheless there are articles with valid criticisms well worth reading. I ran across one recently by Masanoya Sano on Nikkei that asks a good question: ‘Does taking 14 years to deliver Mobile PASMO mean the transit IC card foundation is crumbling? While I don’t agree with everything Sano san says he makes a good case that Japan Transit IC association members are failing in the face of a hydra-headed crisis: declining population with less ridership, competition from other payment services such as PayPay and EMV based VISA Touch, and ridership killing COVID lockdowns. He argues that transit companies must fix some basic problems if the Japan IC Transit standard is to survive:

  • Increase coverage: get all transit on the Transit IC card service map
  • Go mobile: for all transit cards
  • Improve the transit IC card architecture: improve compatibility and loosen up current restrictions for cross region transit, and the ¥20,000 stored fare limit

I believe most, if not all of these can be addressed with next generation FeliCa + 2 in 1 Suica (aka Super Suica) launching this year and deeper payment infrastructure sharing between transit companies. Nothing is guaranteed of course but here’s a look at each category and possible solutions.

Coverage
The transit IC coverage gap is the biggest failure of Japanese transit companies and there are big gaps. Suica only covers major population areas in Tokyo, Niigata and Sendai, roughly half of the stations on JR East are not wired for Suica. A similar situation applies to the other JR Group companies. JR East has promised to get their entire rail network on Suica with a simplified lower cost cloud based Suica in the 2020 fiscal year ending March 2021 but has yet to announce any details (they are specifically referenced in the new Suica Terms and Conditions effective March 27).

On the plus side JR West is expanding ICOCA coverage with a light rail approach of incorporating NFC readers installed in the train car for tap in/tap out for unmanned stations. No wires. SMBC and VISA use the same strategy for their VISA Touch transit boutique marketing program. It’s a practical low cost strategy for lightly traveled rural lines that reduces the hard wire requirement. Only stations that need it get wired and even those installations can use the lower cost JR East cloud based system.

JR West ICOCA area expansion includes on train NFC readers starting March 13, 2020

All major transit companies need to install these lower cost solutions to fill the transit IC gaps and integrate remaining isolated regions. VISA Touch transit boutiques are marketed as a solution for inbound and casual users, but these EMV only installation leave those transit areas off the transit IC grid for regular users and don’t work for wider area travel.

Mobile
Mobile Suica and Mobile PASMO combined represent 80% of the current transit IC card market. Mobile ICOCA (JR West) is due to launch in 2023. There is no word yet about mobile for TOICA (JR Central), manaca (Nagoya City Transit rail/bus), PiTaPa (Kansai region private rail/bus), Kitaca (JR Hokkaido), Sugoca (JR Kyushu), nimoca (Nishitestsu), Hayaken (Fukuoka City Transit). This is a big challenge but the borrowed Suica infrastructure used for Mobile PASMO is a strategy that can be applied to the other major cards.

Improving Transit IC
JR East is releasing the 2 in 1 Suica card architecture that incorporates new FeliCa OS features the most important being the “2 cards in 1” Extended Overlap Service. New regional transit card using this new FeliCa OS and Suica format are launching this month in Aomori, Iwate and Utsunomiya. The next challenge for JR East is expanding 2 in 1 Suica to existing and important region transit cards inside the JR East transit region such Niigata Kotsu Ryuto and Sendai City Transportation Bureau icsca. The JR Group has cooperated to deliver cross region commuter passes which started in

The ultimate long term success of the Japanese Transit IC systems depends on infrastructure sharing and integration. For this to happen other JR Group companies and private rail outside of the JR East regions have to incorporate the 2 in 1 Suica format and improvements for their own cards and regions. Only when all Transit IC Mutual Use Association members are using the new format can they link and combine services in new ways, and add new features such as raising the stored fare card value above the current ¥20,000 limit.

Will it be enough? I have no idea. Immediately I see problems for the Kansai region PiTaPa card association companies (Hankyu, Hanshin, Keihan, Kintetsu, Nankai) as they have to make fundamental changes to use the new card format. I don’t see a Mobile PiTaPA in its current incarnation and this is why SMBC (who run PiTaPa card accounts) and VISA are targeting the Kansai area for VISA Touch transit: non-JR Kansai transit companies have their backs against the wall and no way easy forward to mobile except for going all in with JR West Mobile ICOCA, or taking what SMBC offers them.

Open Loop competition
Kansai area private rail companies never managed to create the equivalent of PASMO. PiTaPa is a postpay card that has credit card issue checks and cannot be purchased at station kiosks like all other transit cards for casual use. Issue is limited, so Kansai transit companies issue JR West ICOCA commuter passes for people who can’t use credit cards. This is the context surrounding the SMBC VISA Touch transit for Nankai announcement that got lots of press attention as the first major test deployment of open loop on a Japan Transit IC card system.

Junya Suzuki’s latest Pay Attention installment has a deep dive on the VISA Touch Japanese open loop transit system solution powered by QUADRAC Q-CORE server technology. It is the solution also used for the Okinawa Yui Rail monorail fare system that integrates Suica/Transit IC and QR support. He argues that open loop EMV is good enough because, (1) we don’t need the over-spec FeliCa 200 millisecond (ms) transaction speed (it’s actually faster, between 100~150 ms), (2) it has a leg up on future MaaS and cloud integration. Holding onto Suica local transaction performance as ‘faster/better’ is a myth holding back progress.

I have tremendous respect for Suzuki san and his work but his arguments fall down for me here. He completely ignores the white elephant in the room: closed loop is here to stay because the open loop model cannot support all fare options. Even on the open loop systems that he champions, Oyster and Opal for example, closed loop cards are still essential and are transitioning to a closed loop EMV model for digital wallet issue. The only change is the closed loop card transition from MIFARE to EMV because bank partners are running the transit system account system backend instead of the transit company. In other words it has nothing to do with technology at all, it is bank system convenience. Bank convenience is what it all boils down to.

Making the right technology choices are essential in our era of limited resources, ride the right horse and you succeed. I want to believe the cloud holds the promise to extend transit IC to low transit volume rural areas that don’t have it now, but every time I use a slow cloud based stera payment terminal I’m reminded how impractical that approach is for stations with high transit volume.

Does it make cost sense to replace the current transit IC system and re-create it with EMV open loop when Opal, Oyster and OMNY systems will always need closed loop cards? The practical thing is leveraging a good system already in use. Upgrade the Japan Transit IC system we have now, spend precious resources that fix current limitations and extend it with new technologies like UWB Touchless.

The strength and weakness of the Japan Transit IC standard is that it’s not top down but based on mutual cooperation. It’s not one entity but association members have to move forward as if they are one. JR East has been the technology leader and is working to improve and share it at lower cost. 2021 is not the make or break year for Japan Transit IC, but it will be an important and challenging one that will set its future direction.

Related post: The 2 in 1 Suica Region Affiliate Card

The good old Japan Transit IC card mutual use map, all the little one way arrows marked with the ‘IC’ logo pointing outside the main IC area indicate transit system compatibility.

iOS 14.5 Face ID impressions

It was just a year ago when iOS 13.5 introduced a small Face ID UI tweak that bypassed Face ID and went straight to the passcode entry screen but over time I did not find it very useful. Face ID took longer and longer to bring up the passcode entry screen, as if it was trying to look past the mask. While it was good that Apple finally acknowledged Face ID shortcomings with face masks after ignoring complaints from Asian users, it took a health crises to force Apple to do something about it. In the end it didn’t change anything.

And now we have iOS 14.5 with Face ID ‘Unlock with Apple Watch’, another stop-gap until Apple delivers a real solution. It will never work with Apple Pay, which it should not though many will wish for it fumbling with iPhone Face ID authorization in the checkout line. It’s probably most helpful when digging for point reward QR Code apps that don’t use Face ID for sign in. Will it help sell Apple Watches? Perhaps, but it also might dampen future iPhone upgrades with improved Face ID.

Some first impressions…it feels like what it is: a clever hack but a hack nevertheless, to do something Face ID wasn’t designed to do that re-routes the ‘chain of trust’ from one way to two way. This makes things much more complicated. Already there are complaints of Face ID unlock with Apple Watch not working when Apple Pay Octopus users are in transit. I also find it unreliable especially during Suica transit. Overall 1 out of 3 times it strikes out. I know the feature is beta 1, but I already get that iPhone X NFC problem vibe: deep down this feature isn’t going to work reliably…ever.

A mystery message when Face ID unlock with Apple Watch doesn’t work