CA$ 93
pledged of CA$ 29,000pledged of CA$ 29,000 goal
5
backers
0seconds to go
Funding Unsuccessful
The project's funding goal was not reached on Thu, July 17 2014 3:12 PM UTC +00:00
Milton FriesenBy Milton Friesen
First created
Milton FriesenBy Milton Friesen
First created
CA$ 93
pledged of CA$ 29,000pledged of CA$ 29,000 goal
5
backers
0seconds to go
Funding Unsuccessful
The project's funding goal was not reached on Thu, July 17 2014 3:12 PM UTC +00:00

About

Thanks for taking time to check out the TRIAT (Tiny Researcher In A Tube) project. For the past 18 months I've been working on developing a working prototype that allows me to fully control a software and hardware platform for collecting GPS data.  Below is a general explanation and then a more technical overview of the device.

General Overview

The motivation for this project is development of a GPS data logging device for a social science research program I would like to carry out that compares how people move around their neighbourhood with the strength of their social ties. 

My hunch is that there is a connection between how much time we spend moving around our neighbourhood or city and how socially connected we are. If this hunch proves  true, it could be a very useful way of understanding the very complex  patterns of our social ties.

I'm not focusing on individuals but at representative groups within specific areas (Census Tract or CT). I will examine the pattern of all the paths collected into one big spaghetti plate of paths and then compare different areas to see if they differ and why. Collecting that data needs a research device that is different than what you can buy off the shelf - TRIAT is that device/platform.

This simple graph may help explain one application for TRIAT, though there are doubtless many others.

Example of research framework
Example of research framework
Early exploration of commercial GPS loggers
Early exploration of commercial GPS loggers

 This was a useful phase and was the early stimulus for developing a device that didn't need to be recharged, that didn't have a bunch of external buttons that could be changed, and where software could be directly accessed for full customization.

Workbench notes and components
Workbench notes and components

Marketplace devices can't be modified and controlled adequately and also require battery charging too often with many features that I don't need. I have pages of notes on commercial products that didn't meet the parameters I needed. In one instance, a commercial research device supplier could provide most of what I needed but each device cost thousands of dollars making it impossible for all but the most heavily funded researchers and explorers to use. 

It is essential that spatial data explorers are able to fully control the features of the device including storage capacity, battery capacity, and how often data is read. Open source components and software provide means that the device can be used either as it is or fully modified depending on the skill level and interest of the researcher.

Device Technical Aspects

I had been learning to write C++ as I used the Arduino open source hardware platform and noted that all of the tools, support and hardware appeared to be in place to make the idea of a fully accessible and modifiable data logging device possible.

Other researchers I knew in human geography used over-the-shelf units and biologists I knew accessed very costly specialty devices. Neither solution met my crowdsource, DIY, collaborative approach to project development and research.

I began by working out how each component worked with the Arduino platform: GPS module, SDmicro card, possible power saving approaches, size, cost, scalability and ideas for cases/housings.

I developed two versions of a working prototype. The first started with an Arduino Uno and an LS20031 GPS module. Using the Arduino sketches I could write data from the GPS module to the Arduino serial monitor.

Arduino Uno and LS20031 GPS module
Arduino Uno and LS20031 GPS module
Bench testing power use of various TRIAT components
Bench testing power use of various TRIAT components

 Power consumption is a key consideration so understanding exactly how each component uses energy and how it can be conserved is critical to the success of TRIAT.

Screen shot of LS20031 data from October 2012 test
Screen shot of LS20031 data from October 2012 test

 The early data gathering experiments helped build a deeper skill set and a better understanding of what the NMEA 0183 data produced by GPS modules looks like and how it can be made more useable.

Running bench trials to check coding and hardware function.
Running bench trials to check coding and hardware function.

 Each new component changes the system. Finding out how the components function in concert with each other is essential.

Example of Data from core set-up without sleep or other functions added.
Example of Data from core set-up without sleep or other functions added.

 Writing data to an SDmicro card is an important step in moving from serial monitor output to independent data logging.

Development workbench
Development workbench

Making careful notes on all the details of development helps to refine the prototype and ensure learning is carried forward at each step.

I explored the many possible dead ends in terms of coding, hardware, production volumes and other factors to ensure that the challenges in moving from prototype to production are solvable. 

I will need funding to work out coding details, develop a user manual, find very energy efficient microprocessor and component combinations, add sleep functions, vibration/motion sensor, or an accelerometer, and ensure the platform is highly reliable. 

TRIAT tube casing.
TRIAT tube casing.

The case will need to be further refined and tested so that it can be readily carried with research participants, doesn't look like other portable devices and is durable. I worked to develop an interesting design that maximized flexibility (length and diameter of acrylic tube can be changed), integrated natural and synthetic materials in an advantageous way, and looked like an interesting scientific object: test tube meets GPS meets corked bottle. Cork and clear acrylic just look cool together.

All of this needs be production ready as researchers will need them in multiples. Some users will not want to build their own and will instead be interested in ordering them in batches. All of the above materials and sub-components are easily sourced and don't require specialized tools.

Cutting acrylic tube for prototypes with mitre box.
Cutting acrylic tube for prototypes with mitre box.
Early prototype with components
Early prototype with components
Notes on adding in different sleep/energy strategies.
Notes on adding in different sleep/energy strategies.

 I was invited to display the TRIAT prototypes and concept at the Subtle Technologies conference in Toronto which was hosted at the Ryerson School of Architecture. It was a great chance to meet other hardware and citizen science developers and see the excellent projects they were involved in. I also gained valuable feedback on the project and device. I was able to ask people directly, "Would you carry this for 7 days?" or "Does this clearly look like a scientific instrument?" and so on. Strong positive responses from people. 

Thank you to James Ruxton (Co-founder and Director) and Farah Yusuf (Curator, Open Access) for the invitation to participate.

TRIAT at Subtle Technologies in Toronto
TRIAT at Subtle Technologies in Toronto
Closer shot of the TRIAT display table
Closer shot of the TRIAT display table

Risks and challenges

Risks with this project include the following:

1. Working out coding issues and glitches will be a challenge. The code needs to integrate sleep functions including an accelerometer or vibration sensor to maximize battery life when there is no movement. SD micro cards provide significant memory but reliability in writing data to a text file that is simple and reliable will be important. I have identified back-up C++ physical computing code support if that becomes necessary.

2. Alternative strategies such as writing the stream of GPS data coming off of the module directly to the SDmicro card could be used. Data cleaning and processing would then be done after collection on a standard computer meaning less processing effort for the TRIAT device.

3. If the volumes of orders are high, the current acrylic tube case design may need to be completely reworked. There are many options that have been explored from small electronics project boxes to 3D printed cases to wearable wrist-watch style cases. Each approach has challenges in terms of cost and production. Some work better for low volumes (3D printed cases) while others are more complex and would only be viable at high production levels (molded wrist-watch style cases).

4. Local production capability for the acrylic tube style cases is sound including commercial level plastics fabrication facilities in the area. There could be pricing and project delivery timing issues depending on the volume of orders. In particular, if productions runs come in uneven batches, smaller orders may end up as a lower priority for the producer.

5. The current prototype requires careful soldering, assembly, and uploading of code. This takes time and requires a specific skills. Finding people who can perform these tasks consistently and at a reasonable cost will be a challenge with higher reward fulfillment numbers.

6. Testing of each device to ensure adequate performance may also be a challenge in terms of time and production. An efficient process will need to be developed for higher volume orders.

7. Customer support for people who are attempting to use the device could require more time than expected. Using an Arduino platform means that there is a very strong community of support but the technical nature of the code and hardware can still be difficult to understand. It may be necessary to develop an online user website where default code can be downloaded in the even that a user changes their code and renders the device inoperable. This would take time and additional resources.

8. Field testing the project and finding unexpected problems with battery life, hardware reliability, or the case could all delay fulfillment of reward orders.

TRIAT Production and Fulfillment Plan

By the end of June, final testing of software and hardware on the prototype will be completed. Ten cases will then be built and field testing will conclude by the end of July. By this time, the funding status of the project will be known.

TRIAT tubes for backers will be produced by the end of July. All components have been sourced, milling and other processes have been tried and will be capable of turning out hundreds of TRIAT tubes if needed. This production can be scaled up further if needed. TRIAT tubes will be mailed to backers by the end of July.

TRIAT prints for backers will be completed by the end of July. These will only be offered on the initial cycle of funding forming a limited run piece. These will be mailed out by the end of July.

TRIAT working units will have all of the components delivered by the end of August. Assembly will take place in August, allowing time for back-ordering of components. All of the components are available from multiple vendors so that sourcing is reliable and has enough redundancy to avoid critical shortages.

Software
GPS logging – TinyGPS as base
Sleep Function – coded into TinyGPS core
Vibration/Accelerometer – coded into TinyGPS core

Hardware
Atmega 328P IC (bare) running code directly
Arduino Pro Mini (alternate)
OpenLog – SparkFun / SDmicro cardholder (alternate)
UP501 GPS module / G635 / LS20031 (alternates)
Batteries – Lithium Thionyl Chloride non-rechargeable
Assembly – conversion of testing bench to small production line
On/Off switch – simple, SPST micro switch
Accelerometer
Piezo vibration sensor – small, weighted, high sensitivity

Cases
Working Cases
– cut, sanded, cork components added, racked, tagged
Souvenir Cases for Backers – same as above

Field Tests
10 Test Units
7 day field trial

Data Review
- How many days logged
- Quality of the returned data
- Conversion to Google Maps
- Failure/Error rate
- Usefulness for network analysis

Hardware Review
- Battery levels
- Component function and condition
- Failure rate

Revisions based on results of above completed by the end of August.
Final fulfillment of working units for major backers completed by the end of September.

Learn about accountability on Kickstarter

Questions about this project? Check out the FAQ

Support

  1. Select this reward

    Pledge CA$ 5 or more About $4

    CHEERING LEVEL: Recognition in report.

    Less
    Estimated delivery
    1 backer
    CA$
    Kickstarter is not a store.

    It's a way to bring creative projects to life.

    Learn more about accountability.
  2. Select this reward

    Pledge CA$ 25 or more About $19

    DESIGN LEVEL: TRIAT Limited Edition Print and recognition in report.

    Less
    Estimated delivery
    Ships to Anywhere in the world
    3 backers
    CA$
    Kickstarter is not a store.

    It's a way to bring creative projects to life.

    Learn more about accountability.
  3. Select this reward

    Pledge CA$ 25 or more About $19

    MOMENTO LEVEL: TRIAT souvenir tube with logo and recognition in report

    Less
    Estimated delivery
    Ships to Anywhere in the world
    0 backers
    CA$
    Kickstarter is not a store.

    It's a way to bring creative projects to life.

    Learn more about accountability.
  4. Select this reward

    Pledge CA$ 150 or more About $113

    INVENTOR LEVEL: Working device with instructions and recognition in report.

    Less
    Estimated delivery
    Ships to Anywhere in the world
    0 backers
    CA$
    Kickstarter is not a store.

    It's a way to bring creative projects to life.

    Learn more about accountability.

Funding period

- (30 days)