Symbian Educating its Developers to Location Based Services

In February 2004, John Pagonis and Jonathan Dixon of Symbian (OS for smartphones) published a very refreshing article to educate their developers’ community to Location Based Services. They briefly explain context awareness, location awareness and location based services. The perspective is clear:

What is of interest is what we can do with location awareness rather than how much awareness is implemented or delivered to the device. [...] Unfortunately when talking about LBS, most developers and users think in terms of static user-initiated scenarios like finding the nearest taxi rank or getting a map proximate to the user’s position.

and they even go into mentioning manual positioning

Many times it suffices for a location awareness application to be useful for the user to manually inputs its approximate location to the device

They clearly explain the difference between accuracy and precision

All quantitative properties provided by location acquisition technologies have a realted accuracy (or error or tolerence) and precisions (of confidence). Coarsely, the accuracy is a mesure of how far off the data may be and the precision is what statistical percentage of the time the actual (or measured) error will be within the accuracy estimate. These are a function of the acquisition technology, and current (and maybe past) operating conditions. Qualitative information can only have a confidence.

and mention a “graph of confidence over accuracy”

In actual fact, this will be a graph of confidence over accuracy, forming some sort of curve, such as a Gaussian distribution, that will itself be a function of the acquisition technique is use.

They also come up with interesting insights and conceptsincluding:

• “Local positioning” which make use of short range networks
• We should not expect the support of E911 and EU E112 to be exposed to developers
• Human-Device-Network interactions
• Constraints include power consumption, cost, autonomy, speed, networking, autonomy, processing and overall system complexity.

Unfortunately, part II that is supposed to talk about application scenarios is nowhere to be found and maybe was not ever written…