21 April 2010

Designing the Internet of Things

Filed under: Internet of Things, mashup* event, Mobile Monday — David Wood @ 2:10 pm
  • Computers; smartphones; … smart things.
  • The Internet; the mobile Internet; … the Internet of Things.

These two epic trends tell aspects of the same grand story.  First, computing power is becoming more widespread, more affordable, more compact, and more miniature.  Second, networked intelligence is becoming more widespread, more affordable, more effective, and more informed.

As a result, we can look forward to a time, in just a few years, where each of us owns more than a dozen different devices that communicate with each other, wirelessly and transparently.  That will take the number of wireless modems in use in the world to upwards of 50 billion.

Going further, there are forecasts of no fewer than one trillion wirelessly connected “things”, where this time the connection will often involve simpler connectivity such as RFID.  As reported recently in Wireless Week:

There will be 1 trillion devices connected to the Internet by 2013, said Cisco Chief Technology Officer Padma Warrior during her Wednesday keynote address at CTIA.

Warrior argued the boom in connected devices, applications and mobile broadband would change not only the wireless industry but society in general.

“The Internet is no longer just an information superhighway, it’s a platform,” Warrior said, citing the increased adoption of M2M technologies and the exponential growth of apps…

To prove her point, Warrior moved through a series of technology demonstrations with a Cisco colleague that detailed what it would be like to interact with next-generation mobile technology.

The pair showed off augmented reality in a subway system; location-based advertising and mobile coupons; and a mobile telepresence app.

The next big revolution that will happen is the Internet of things,” Warrior said…

Evidently, the ever lower costs and increased quality of computing and connectivity are opening all kinds of new opportunities.  It’s easy to speculate on possibilities:

  • Distributed arrays of sensors that can more reliably – and more quickly – highlight the changing concentrations of volcanic ash;
  • Luggage tags that know (and can report) where your luggage is;
  • Air conditioning units and heating units that can coordinate to act in concert, rather than independently;
  • A handheld toothbrush that can let you know if you’re not putting enough effort into cleaning the innersides of your lower right molars;
  • Smart sticking plasters that detect microscopic changes in skin condition or blood flow;
  • A monitor that can detect if you are too distracted (or too dozy) to drive safely.  (Even better, put the driving intelligence into the car itself, rather than rely on human drivers.)
  • Surveillance cameras that can analyse what they are filming, being alert for security abnormalities;
  • Audio recording devices that can understand what they are hearing;
  • Smart glasses that can interpret what you’re looking at;
  • Smart digital signs that change their display depending on who’s looking at them;
  • And all of these devices connected together…

Does this sound good to you? We can debate some of the points, but overall, it’s clear this grand technology trend has great power to improve health, education, transport, the environment, and human experience generally.

So why isn’t it happening faster? Why is it that, as an industry colleague said to me recently, this whole field is in a state of chaos?

It’s chaotic, because we don’t know what’s happening next, nor how fast it’s happening.  The use cases developers identify as important at the start of a project often turn out to be less significant than ones that turn up, unexpected, part way through the project.  (That’s not necessarily a problem.  It is, of course, a large opportunity.)

In short, although lots of the underlying technology is mature, the emerging industry of the “Internet of Things” is still far from mature.  The roadmaps of product development remain tentative and sketchy.  Speedy progress will depend on:

  • A few underlying technology issues – such as network interoperability, smart distribution of tasks across multiple processors, power management, power harvesting, and security;
  • Some pressing business model issues – since not all existing players are excited by the prospects of cost-savings which would in turn reduce their profits from existing products;
  • Some ecosystem management issues – to solve “chicken and egg” scenarios where multiple parts of a compound solution all need to be in place, before the full benefits can be realised;
  • Some project development agility issues – to avoid wasted investment in cases where project goals change part way through, due to the uncertain territory being navigated;
  • Some significant design issues – to ensure that the resulting products can be widely accepted and embraced by “normal people” (as opposed just to early adopter technology enthusiasts).

These are some of the themes that I will seek to explore as one of the speakers at the mashup* event in London on Tuesday 4th May, entitled “Internet of things: Rise of the machines“.  In addition to myself, the other announced speakers are:

Many thanks to the mashup* team for organising this get-together!  I hope to see you there.

Footnote: The ReadWriteWeb publish a useful series of articles about the Internet of Things.  At time of writing, the most recent article in this series is “Internet of Things Can Make Us Human Again“, which highlights the ideas and work of David Orban, Founder of WideTag.  Here’s a brief extract:

Orban’s dream is that thousands of years of human subservience to machines will end because we will teach our machines how to not only take care of themselves, but how to take care of us as well…

These new machine networks will be so redundant and reliable that we will be freed from most of our machine-operating duties. We will get to be human again…

Extending these ideas, David recently spoke on the theme “Free to be human” at Mobile Monday Amsterdam.  It’s a great introduction to many Internet of Things ideas:

By good fortune, David will be in London this weekend, since he’s one of the speakers at the Humanity+ UK2010 event.  He’ll be addressing the Internet of Things once again, along with some thoughts on the progress of the Singularity University.  In my experience, he’s a fascinating person to talk with.

29 January 2010

A strategy for mobile app development

Filed under: Agile, applications, consulting, fragmentation, mashup* event, mobile web — David Wood @ 12:15 am

The mashup* event in London’s Canary Wharf district yesterday evening – hosted by Ogilvy – addressed the question,

  • Apps: What’s your strategy?

The meeting was described as follows:

This event will help people in strategic marcomms roles understand the key challenges with respect to apps and identify the building blocks of an app strategy:

  • What are the platform choices?
  • What are the app store choices?
  • What devices should you support? …

mashup* is bringing together several industry experts and specialist developers to help demystify, clarify and explain the issues around the rapidly emerging Apps channel…

The event was sold out, and the room was packed.  I didn’t hear anyone question the need for companies to have a mobile strategy.  Nowadays, that seems to be taken for granted.  The hard bit is to work out what the strategy should be.

One of the speakers, Charles Weir of Penrillian, gave a stark assessment of the difficulty in writing mobile apps:

  • For wide coverage of different devices, several different programming systems need to be used – apart from those (relatively few) cases where the functionality of the app can be delivered via web technology;
  • Rather than the number of different mobile platforms decreasing, the number is actually increasing: fragmentation is getting worse;
  • Examples of relatively new mobile platforms include Samsung’s bada and Nokia’s Maemo.

One mobile strategy is to focus on just one platform – such as the Apple iPhone.  Another strategy is to prioritise web-based delivery – as followed by another speaker, Mark Curtis, for the highly-successful Flirtomatic app.  But these restrictions may be unacceptable to companies who:

  • Want to reach a larger number of users (who use different devices);
  • Want to include richer functionality in their app than can be delivered via standard mobile browsers.

So what are the alternatives?

If anything, the development situation is even more complex than Charles described it:

  • Mobile web browsing suffers from its own fragmentation – with different versions of web browsers being used on different devices, and with different widget extensions;
  • Individual mobile platforms can have multiple UI families;
  • Different versions of a single mobile platform may be incompatible with each other

The mobile industry is aware of these problems, and is pursing solutions on multiple fronts – including improved developer tools, improved intermediate platforms, and improved management of compatibility.  For example, there is considerable hope that HTML 5.0 will be widely adopted as a standard.  However, at the same time as solutions are found, new incompatibilities arise too – typically for new areas of mobile functionality.

The suggestion I raised from the floor during the meeting is that companies ought in general to avoid squaring up to this fragmentation.  Instead, they should engage partners who specialise in handling this fragmentation on behalf of clients.  Fragmentation is a hard problem, which won’t disappear any time soon.  Worse, as I said, the nature of the fragmentation changes fairly rapidly.  So let this problem be handled by expert mobile professional services companies.

This can be viewed as a kind of “mobile apps as a service”.

These professional services companies could provide, not only the technical solutions for a number of platforms, but also up-to-date impartial advice on which platforms ought to be prioritised.  Happily, the number of these mobile-savvy professional services companies (both large and small) is continuing to grow.

My suggestion received broad general support from the panel of speakers, but with one important twist.  Being a big fan of agile development, I fully accept this twist:

  • The specification of successful applications is rarely fixed in advance;
  • Instead, it ought to evolve in the light of users’ experience with early releases;
  • The specification will therefore improve as the project unfolds.

This strongly argues against any hands-off outsourcing of mobile app development to the professional services company.  Instead, the professional services company should operate in close conjunction with the domain experts in the original company.  That’s a mobile application strategy that makes good sense.

Blog at WordPress.com.