(I’m writing these as an exercise in thinking out loud and starting some dialog. I make no claims of breaking new ground here and welcome criticism. –jet)
The concept of a spime is simple to understand, the path to a world of spimes is not as obvious. Instead of attempting to invent spimes out of whole cloth we can extrapolate from existing capabilities and look to what we can do now or in the immediate future. Between the spime-less now and the spime-full future we will go through a brief period of working with what I’ll call “proto-spimes”.
Proto-spime: an existing object retrofitted for spime-like behavior by tacking on some number of sensors, computing resources and communication mechanisms. A proto-spime exhibits spime-like behavior but is no more a spime than a person who speaks another culture’s language is a member of that culture.
The arbitrary line I draw between a proto-spime and a spime is that of design intent. A proto-spime was not intended to have spimelike behavior when it was initially conceived and designed; a real spime has intent in the initial conception and design. Compare this to early portable personal computers and modern laptops: Early portable computers were PC-ATs smushed into portable cases while modern laptops are not only designed and built on the plan of portability but often contain features unique to portable devices or lack those found in non-portable devices.
The process and benefits of retrofitting items with simple magnetic or radio resonant tracking tags is well understood across a number of industries. Inventory management systems can easily track items that have a small barcode on the surface or a resonant tag embedded inside; these range from books leaving a library to a rental car being returned. This model of items being managed by users is one starting point for thinking about proto-spimes. The requirements and processes are well understood, we’re simply extending and enhancing an existing model with smarter tags.
Making a proto-spime
The resources needed for spimelike behavior — basic data collection, storage and retransmission– can currently be packaged in a small space and had for a relatively low cost in volume. Maxim’s iButtons are a current example of this sort of device. About the size of a few stacked dimes, iButtons can collect or store data and can draw their operational power from their communication channel. (The 1-Wire protocol is also used by personal weather stations and other devices needing minimal operational power for data collection and communication.)
Spime Retrofit Module (SRM): a device capable of data collection and transmission that does not communicate with or control its host. A SRM possibly has limited decision-making capability in the domain of what data is collected and who it is shared with.
Let’s assume for the sake of design that our basic spime retrofit module is larger than an iButton, about the size of a 9v battery. Let’s also assume that a fancier version with limited computing power and the ability to initiate communication with other spimes is the size of a two or more D cell batteries. The basic SRM can log limited amounts of data and transmit it on demand from a powered reader; a fancy SRM can collect data, make simple decisions, initiate contact with other spimes, or decide when to broadcast data to passive sensors.
Batteries are good models for our SRM for a few reasons: the size and weight are understood across most cultures, they are easily fitted into a variety of existing items, and most importantly, there are plenty of CAD models already in existence.
So what do we do with these spime retrofit modules?
Initially, SRM’s can be easily attached to or installed in existing items that their humans want to know more about (or will soon discover they want to know more about). Some of these items might not be worth redesigning as proper spimes while others might be more than useful with an embedded SRM.
Once we’ve learned a few lessons with proto-spimes we’ll be able to include the other side of spimes — data collection and management — in the iterative development process of spimes and SRMs. For now we’ll just stick to uses of the first generation of SRMs. (With the addition of the ability to directly manipulate its physical environment, a SRM becomes an autonomous, data collecting robot. I think these are interesting but because they can manipulate their environment they don’t match my definition of a proto-spime as a passive data collection device.)
Early Adopters of Proto-Spimes
The obvious adopters of proto-spimes are organizations that manage, move and maintain inventory and items. There are other early adopters in areas where individuals or groups have an interest in data and metrics collection to use as feedback for performance or entertainment purposes.
People responsible for equipment used by others
IT departments will be able to track the physical history of laptops, portable projectors, or other high-value and easily-damaged items. How often does the new high-end projector actually get used and in what conference rooms? Why does it always spend the weekends at the same person’s house? How and where do laptops get broken? Where does the tool cart spend its time during the day and what sort of equipment is it used to move?
Tool and equipment rental firms can verify that their rental items weren’t subjected to environments that violate rental terms or safety restrictions. In addition, they’ll be able to know how long a tool was used and in what ways, information critical for both routine maintenance and product marketing.
Auto rental companies and other transportation firms are already using proto-spimes in the form of black boxes that collect driving data and track the location of rental cars.
Athletes and Coaches
Data and metrics collection are a key component of most sports and many athletes and coaches are already moving into proto-spime territory to gain a competitive edge. This is already common in motor-sports, vehicles are heavily instrumented and the data is relayed in real-time to team managers and engineers. Other types of sporting activity has yet to directly benefit from this sort of real-time data collection and analysis.
Cyclists currently use integrated computers that collect information on things like heart rate, pedaling cadence, altitude, and speed. These are displayed in realtime for feedback to the cyclist while riding and saved for later post-ride analysis. With some simple (and hopefully lightweight) additions, these proto-proto-spimes will not only collect data from their bike but share data with nearby cyclists or even spectators. Teammates will be able to use shared data to optimize their pack performance while riding and more easily anticipate needed changes in ride order or speed based on one another’s physical performance. (All I know from bicycle racing is listening to my friends talk about it. Feel free to fill me in on the reality.)
Sports that use objects — hockey, football, basketball, golf — will be able to collect information from the objects themselves during and after an event. How hard a ball was kicked, how damp the green was, the precise amount of time between events and information on the playing surface and environment will all be immediately available to interested parties. Players of team sports that use objects can collect data on the performance of other players by monitoring objects to look for patterns of weakness, strength, or exhaustion that are not normally perceptible during game play.
Environmentalists, Anthropologists, and Other Observers
SRMs are drop-in replacements for many of the tracking devices used by people who monitor objects in the environment. Instead of simply relaying position, they will be able to collect information on their environment, save that information and exchange it with other SRMs in the wild. SRMs will be able to feed information to one another on additional data that needs to be collected, back up data carried by other SRMs in order to ensure that data gets back to the SRM’s owner.
I think this might be the most important area for design and development of proto-spimes and spimes. The ability to collect and dynamically manage data from the field and ensure a high return rate of that data to the collector is useful for everything from weather forecasting to studying the remote corners of the planet.
An entire pod of dolphins could be instrumented for data collection and return not only information about their world, but information about the dolphins who did not return and why they did not return.
An SRM that is small enough and sturdy enough to survive the digestive process might go from an organism low on the food chain to a higher organism. While it may never be retrieved, if it senses a friendly/compatible SRM nearby it could relay its history and increase the chances of its data getting back to the owner
The handful of examples here has me thinking about a number of engineering, legal and ethical questions:
- Who owns data collectd by a spime?
- What are the rules of engagement for spimes in the wild?
- What data can I legally collect and store?
- How do I protect my data?
- How do I protect my spimes from being spimejacked? (And am I the first perso to use the work “spimejacked”?)
- Are self-transporting spimes useful? How are they different from “software agents”? Is it time to dust off my Telescript developer’s kit?
- How do we prevent ill-effects from all the RF fields used by spimes, both physical damage to the environment and harmful interference to other users of the airways?
- How do we make SRMs that are low-impact on the environment should they become lost or damaged?