Back to the Future - Morse Code and Cellular Phones

by Brian McConnell

I've spent most of the past five or so years thinking about handheld devices, their limitations and how to work around them. Having worked with telephones since I was in high school, this has been something of an obsession.

The hot trend today is to cram every feature imaginable into mobile telephone handsets. This has led to some cool things like camera phones, mobile gaming, and such. The problem is that a lot of designers overlook some basic limitations in these devices, and more importantly, the situations in which people use them.

Cellular phones are all about mobility. Good mobility applications recognize that the user is often in motion (walking, driving, etc). Safety and convenience require that the application should demand as little visual attention as possible. Badly designed applications force the user to stare at the telephone's display instead of paying attention to surrounding environs. This is why speech user interfaces work so well for mobile users. They allow the user to interact with a service in a "heads up" stance, without looking at the phone. Unfortunately, most mobile applications are of the badly designed "let's take a PC interface and shrink it down" category.

Text messaging is an enormously popular service, but it too suffers from this basic user interface conflict. Sending and receiving text messages requires the user to look at the display. Receiving messages can be done at a glance, so this is not such a burden. Sending them is another story. Some people are adept at tapping messages on numeric keypads, but doing so requires the user to pay attention to the display. Try writing a text message without looking at the phone. Not easy.


Morse Code, or a derivative of it, could be one way to solve this problem. With Morse Code, one could tap text messages out without looking at the telephone, and without having to fumble with ever smaller keypads. I'll admit that the idea of resurrecting Morse Code seems improbable, but then it's worth remembering that only a few years ago, the idea of people typing with their thumbs also seemed absurd.

How might Morse be incorporated into a telephone handset. I sketched out a fairly simple interface. Here's what I came up with.

The telephone would have a fairly large pressure sensitive panel on its back side, big enough that you would not have to look at the phone to locate it. It might also be possible to use the telephone's existing microphone to sense taps (although discriminating between short and long pulses could be a problem).

You'd send messages in a couple of different ways depending on how you were carrying the phone at the time. I devised a couple of tweaks to make the process of sending messages faster.

When carrying the phone at your side, you could send messages with one hand by tapping on the back of the phone in the convention dot (short) and dash (notation). The panel would interpret a brief pulse as a dot, a longer pulse as a dash. Timing is important, so this method of sending messages takes more practice.

With both hands free or with the phone resting on a surface, you could use a slightly different method to tap messages. Holding the phone in one hand and tapping with the other, you'd tap the panel with your fingernail to send a dot, and with your whole fingertip to send a dash. Timing is much less important here, so this method will be easier for people to learn.

Receiving messages is less of an issue, since they'll arrive as text messages. The sending telephone will convert the tapped dots and dashes into alphanumeric messages to be sent via SMS or IP. The receiving telephone will display these in the usual way (an option to play messages via text to speech synthesis would be a nice add-on, and as mobile phones become more powerful, should be easy enough to do).

Hands-Free Mobile Phone Features

Incorporating a Morse Code key into the back of a telephone handset has other uses besides tapping text messages. One of the things this enables you to do is to make it easier to control a telephone in hands-free mode.

For example, you could design the phone so that it recognizes certain codes as keypad commands, primarily for deciding how to deal with incoming calls.

.. = answer call
... = send call to voice mail
.... = forward call to preprogrammed number

So while you're driving along, you could dispatch incoming calls as desired by tapping on the back of the handset, something you could do heads up, without taking your eyes off the road.

While this isn't Morse Code per se, it's the same idea, and it should be easy to train users to learn a handful of short two or three digit codes as in the example above. This is probably more realistic than training users to compose SMS messages in Morse, as anybody can memorize a handful of tap sequences.

Back to the Future

I'll admit this may seem like a bit dated, but even with a Treo 600, I find it difficult to type text messages. It seems to me that something like this is worth a try. The cost of embedding this in a handset should be pretty minimal compared to that of other features like digital cameras. You're basically talking about a small plate attached to a piezo-electric sensor, which is about as simple as it gets. Even better if you can make this work using a phone's existing microphone to sense taps.

Would people actually use this? I don't know. It's hard to tell what will catch on. I thought ringtones and camera phones were improbable at best, and now those are both billion dollar industries. If something like this makes it easier to use SMS, then my guess is that it will catch on, at least with a subset of users.

While the Morse Code application may not catch on outside a small group of power users, the idea of using Morse-like code to control a telephone in hands-free mode makes a lot of sense. Tap twice to answer a call while driving, three times to send it to voice mail, four times to forward the call to your secretary. That'll be easier that opening the phone and pushing a key while driving, and a heck of a lot safer.


2005-06-28 12:46:36
Morse Code-a great idea
I like your idea of using Morse Code for sending text messages. I don't text message myself, in part because of the hassle of finding the right little key for a letter. Perhaps someone could develop some software that could be added to some cell phones. If they did it right and made it open source, some cellular makers might add it as a free feature.

Also keep in mind that learning Morse well enough to send it is much easy than receiving--so much so that when I took my FCC code test for an amateur license, they didn't even bother to test sending. If I could receive 13 WPM, that was enough. So a system requiring the user to know Morse well enough to send, but not have to receive should work fine.

The reason is that when you're receiving your mind doesn't get a break, the letters just keep coming. If you can't recall what letter something is, you quickly fall behind and get flustered. But when you're sending that's not a problem. If it takes longer to recall that v is di-di-di-dah than to remember that e is dit, it's no big deal. You control the timing.

Also, there is some ham software that knows how to translate dot-dashes to characters. Some might be available for this use.

Last but not least, it might be clever to leave open the option of passing along the di-dah sound to a cell phone with earphones. It'd make a great game for kids. Sending di-dah-dit, would mean turn right ahead. Two people could communicate without appearing to do so.

--Mike Perry KE7NV

2005-06-28 13:52:12
Interesting, but...
A couple of thoughts crossed my mind while reading this article.

First, it's difficult enough to remember the right keys to press for T9 text input, and I have been using those key/letter combinations at least weekly for a few years. Successfully using a Morse-like system requires memorizing at least 36 alphanumeric codes, plus punctuation and control codes. I don't think that's going to fly with any but the most dedicated users.

Second, the addition of a timing component introduces an extra dimension for errors. Let's say you want to trigger a triple-tap response (such as forwarding a call in your example). Tap-tap-bump (oops, something hit your hand before you could hit the third tap). By the time you issue the third tap, the phone has concluded you are done tapping, and hangs up instead. You might be able to get around this by making the sequence more unique (one to three taps, followed by a closing dash), but this won't solve the related problem of interpreting tap-touch-tap-dash. If you're imprecise in your tapping, will you know if each tap was registered? Audio feedback, perhaps?

Anyway, an interesting set of ideas. Maybe someone can hack a proof-of-concept into an existing phone, using a side switch perhaps.

2005-06-28 18:02:09
How about T9
Great idea, I certainly would use it, but how about incorporating it with T9 prediction, that way you only have to learn one code for each of the 9 numbers, and these could also be used for entering numbers as well as letters, with the addition of a four-way pad + 2 softkeys for regular navigation, the keypad would be tiny, yet more functional than ever.
2005-06-29 03:52:24
How is that supposed to work? You can’t know T9’s guess at what you’re trying to write without looking at the display – and that’s just what the idea was trying to avoid forcing you to do in the first place.
2005-06-29 04:20:23
Rocker switch as input device?
Quite a few phones currently have
a rocker switch for volume control.

This could be used for 'dots' and 'dashes'
and has the advantage of positive action
without a timing problem. (Short delay,
or another button could be used for end of
word, unless the morse could be reconfigured
to use a terminating encoding.)

2005-06-29 04:52:41
2 finger morsing
I have once seen a morse switch that could be operated with two fingers: one switch for long and one for short. This allowed one to concentrate on the sequence of long and short signals instead of taking care of the right length. Unfortunately, I could not find any information on such a device on the net. Maybe someone of you does.
BTW: I would immediately use Morse code on my mobil. I even switched to Dvorak layout on my PC.
2005-06-29 05:28:35
It's your brain, not your eyes...
I take exception to the idea that a cell-phone entry method that doesn't require visual contact (or even hand contact) with the phone is safer while driving than other methods. If your mind is somewhere else, your brain won't process information coming in: i.e, "the gates of perception."

I was out walking with my toddler on my shoulders the other day when we were about to cross a busy intersection. We saw other cars stopping at an intersection, but there was a women talking on her cell phone who wasn't stopping. I had long enough to say "what an idiot" before she crashed into a boat that was being towed by a pickup truck.

Well, I pick up my cell and dial 911 and it won't connect me until the cheap-ass GPS receiver acquires the GPS signal, and that just takes forever. Fortunately a fire truck was driving down the street, and the firefighters noticed the accident before my GPS receiver noticed the satellites, so I aborted the call.

Even with my hands-free kit I don't feel safe walking at high speeds in heavy traffic, never mind cycling or driving. Just say no to fumbling with gadgets in your car.

2005-06-29 06:23:47
Morse code cellphone vibration
And don't forget the vibrating alert feature on cellphones-- why not have text messages "read" to you in morse code through the vibrating alert? That way you can actually read and receive text messages without having to look OR hear them! Imagine in a meeting, your pocket starts to vibrate... you begin to decode the message, it's "abort abort abort" -- your able to make up an excuse and get out of there before disaster strikes, noone the wiser to your received message.

Morse code messages through vibration alerts is already something used to enhance accessability for handicapped folks, but not something I've seen on a cellphone before. Its kindof an extension of what you were talking about so I figured I'd post about it here :)


2005-06-29 06:31:56
Morse Code for Nokia is already here!
Morse Code for Nokia phones using symbian OS is already here! Check it out!
2005-06-29 06:53:45
2 finger morsing
These are actually very common, a a href="">Google for "morse paddles" brings up many pages, and as soon as you look at the pictures you'll < href="">find them. It looks like in the US these types of Morse paddles are called "bugs", but I'm sure that in the UK they used to be referred to as di- or bi-something.
2005-06-29 07:49:35
A backward look at an experiment
I am probably guilty of sleeping on my chance to popularize this -- I had a Morse Chat application working in 2002 on Danger Hip-tops. The design of the Danger did not permit integration with the SMS stack, so my prototype was limited to a proprietary character-buffered text chat paradigm akin to Morse telegraphy. It was certainly part of my goal to extend this to message-buffered communications such as SMS and IM.

The important thing to realize, of course, is that since the application decoded the text from the input pulses, the media is primarily textual and any subset of users wishing to regard it as a Morse channel are not prejudicing others who may want to see it as a conventional text chat or an SMS exchange.

There is some data on the project here.

My prototype had many of the features outlined above:

1. single-key entry of text
2. machine decoding of input permitting receiving apps (or users) to read the text if this was preferable
3. alternatively, output could be sampled as authentic telegraph dits and dahs, as vibrations, or as LED pulses. In these modes, the timings used to encode the text were retained (to preserve the "fist" of the sender), and options permitted the speed of sending to be altered to suit the copying speed of the listener. In this way, the channels were skill-indifferent.

I should have stuck with it further or moved it to more popular phones. The project was fun and squarely confronted the often ignored fact that the status quo in phone text entry is indeed not very elegant. This one had rough edges, of course, but it was not wholly inferior and indeed had a geek-ready appeal that is perhaps yet to be fully realized.

2005-06-29 09:37:34
Funny, I just mused about this topic last week on my blog:

Quick mind interface thoughts - 0xDECAFBAD

I think others have covered this in previous comments, but you might want to check out the Palm Radio Mini-Paddle:

The other notion I had before discovering this paddle was, instead of one button for taps, two buttons for dits and dahs. I think someone else earlier mentioned hijacking the volume rocker for this. I'd just like to have a couple of spots on the thigh of my pants on which to rap out a quick tattoo, along with some subtle vibration for confirmation.

2005-06-29 10:12:54
Audible feedback for Morse entry
One of the problems with entering Morse as suggested is the precise timing required to indicate the different spacing between characters and words. I have a programable Morse keyer which allows you to pre-record messages. You to enter one word at a time. Each time you finish a word, the keyer goes beep-beep. If you make a mistake, you can erase it by sending 6-or-more dits. The keyer then plays back the previous word and lets you re-enter the next word. For folks who want more precise feedback than beep-beep, you could program the device to actually echo the word just entered.
2005-06-29 10:23:39
Palm as an example
If people are (were?) willing to learn graffiti to talk to their palm devices, I think it's not too much of a stretch to suggest people will be willing to learn Morse to SMS on their phones more quickly.
2005-06-29 10:24:02
Finger Tip Tapper
I have the Palm mini paddle mentioned by a previous poster, and love it (it's built into Shacktopus and can just barely be seen in the lower right corner of the photo here). It retracts, which is great for a portable application, but it's hard to visualize it working with a wee phone. Perhaps more interesting in this context is the approach taken by the Finger Tip Tapper, which still works as an iambic keyer but is flat. I have not tried it, but the reveiwers on seem to like it.

The iambic approach is great for Morse Code, and is very fast once you get used to it. Basically, holding one key generates a string of dits; holding the other generates a string of dahs; and hlding both generates a continuous didahdidahdidah... requiring far fewer physical motions to generate characters. Of course, I would not be one to suggest that people do this while driving (though once it becomes second nature, it really is easy).

Cheers (and 73),
Steve N4RVE

2005-06-29 10:33:02
probably easier than you think
once upon a time, when computers were guys with slide rules and log books, morse code wasn't so hard. in fact when i got my radio license, at 12, sending & receiving was part of the test. i think people who grew up with nintendo controllers welded to their hands could learn morse faster than you might imagine. consider the letter N sms on phones is two keystrokes and the letter S is four. the audible feedback already exists (unless it's been disabled on the phone). for me three keys is enough. a dit on the left, a space in the middle, and a dot on the right. a space between words (duh) and two between sentences. with a lighter-touch keypad you'd be off and running.
2005-06-30 06:18:40
Excuse me - Are you in my way?
This is one of those typical gen BS type posts that are becoming so common these days where the cell phone is decried as absolutely essential equipment for even the 30 second walk to the toilet.
Can we not, just on occasion do one thing at a time? Is there really a benefit to clogging everyone else's senses with your useless diatribe of how you just stepped on a sidewalk crack with your new boots that your boyfriend thinks makes you look soooo good in, that he wiped your chin last week, and oh my god I think he likes me ..... ad Clausius. Please, just shut up and walk to class.
Even though you get some new way of unobtrusively tapping out a message, it just makes the messaging geeks just that much more noticeable and therefore distracting and annoying. I love having to weave in between these useless yahoos that because they're so engrossed in their conversation or reading the useless drivel on their cell phone, that they can't be bothered to look around and realize that they're in everybody's way. I've knocked a few of them on their ass just because they don't have the consideration to look where they're going.
It's not the input method that needs to be redefined, it's the idea of appropriate behavior that needs to be addressed!