Workshop Wrapup: Technology on the Trail Day 2

The second day of the Technology on the Trail workshop at Virginia Tech consisted of a pair of work sessions and a workshop wrapup.

The first work session, led by Nicholas Polys and featuring John Munsell and John Jelesko, looked at science on the trail. It delved into the challenges of taking technology outdoors, balanced with the opportunities that it provides. Of particular concern are problems of cleaning up “dirty data” from erroneous readings. It’s great to get more people involved in data collection, but without knowledge, training, and high-quality equipment, we run the risk of collecting erroneous data.

The second work session, led by the project research associate Grace Fields, focused on her cultural probes. We got to try out some of her “would you rather” probe questions, e.g., would you rather hike on a rainy 60 degree day or a sunny 30 degree day. It was noted that these aren’t opposites (they aren’t meant to be!) and often the answer is “both”. Other probes and, importantly, some early probe results were presented. The results really drove some interesting conversations, and also highlighted the need for follow-up interviews or focus groups to delve deeper into the “why” behind the responses. Alan Dix noted that probes are better at putting forth questions rather than answering them, making it important to discover the key questions that emerge from looking at the probes.

The wrapup sought to both look back as well as look forward. There were great ideas shared about possible partnerships, follow-up events, opportunities for funding, and venues for writing. At this stage of the initiative, it is important to cast a wide net and to work in directions that meet real needs for people and organizations that care about trails and that see value in technology. All are encouraged to share ideas and help out!

2017-03-04 10.03.37.jpg
The Cascades (a bit frosty around the edges)
2017-03-04-10-17-17
Chewbacca (Norm) and Yoda (Scott) staying warm
2017-03-04 10.06.56.jpg
Steve Harrison offering date and fig cake to Alan Dix and the masses

As a quick addendum and final photos: Day 3 saw us match our efforts to our talk, as we hit the trail for a hike to the Cascades. Ten of us made the 4-mile walk in below freezing temperatures to view the iconic waterfall and continue our conversations.

Advertisements

Reading Group Summary: Bridging Urban and Woods Technology

 Readings:

With this week’s paper we examine the connections between urban and trail technology use, considering how lessons from sensor-based science the woods affect urban sensing efforts.

  • Dana Cuff, Mark Hansen, and Jerry Kang. 2008. Urban sensing: Out of the woods. Communications of the ACM 51, 3 (March 2008), 24-33.

Agenda:

  • Brief revisit of reading group and introductions of anyone new
    • Attendance: 4 people (1 professor, 2 graduate students, 1 undergraduate student)
  • Summarize papers
  • Discuss papers

Discussion:

This paper, authored by an architect/urban planner, a statistician, and a lawyer, outlined challenges regarding sensing efforts in urban areas.  They began by describing successes in sensor-based science in the woods, where air, water, soil, etc. sensors could collect data continually without objection from the birds or worms.  They contrasted that with efforts to do sensing in urban areas, where there are legal issues (people object to being sensed more than birds) and a greater likelihood of junk data and garbage analysis.

The issues from this paper resonated with the people in the discussion group, particularly in reflecting on past papers.  Several people noted parallels with the challenges encountered with the Dix data, for which even a careful scientist and a conscientious science community generated faulty data (e.g., GPS malfunctions, forgetting to turn on or off a sensor) or questionable analysis (e.g., is there a “happy” day, or is that the wrong unit of time to examine).  Challenges stemming from large data sets, including human-generated data, certainly have grown in importance in the 8 years since this paper appeared.

The calls for action certainly seem prescient today.  The call for more data commons efforts is seen in data sets like the Dix data, government datasets, and the NSF requirement that all proposals include a data plan that explains how data generated in the proposed work would be made available for others to analyze.  The call for distributed citizen-initiated sensing is seen in efforts like Google traffic (Waves), Foursquare, bar tracking, Facebook check-ins, openstreet map, CMU bus schedule project (Zimmerman), and lots of other crowdsourcing efforts.

And it was encouraging that the authors noted that user interfaces are both important and hard.  Again, we saw that in the Dix data, which was made available but was reformatted  and cleaned up multiple times–and gaining insights from it is still hard: McCrickard’s class had difficulty identifying interesting findings for a week-long homework assignment. 

There were lots of ways forward that emerged from the discussion and paper.  Many were related to the difficulty in knowing the right question to ask about data sets?  A bottom up examination reveals trends and top down reveals questions, but there’s a danger in just finding what you’re looking for without a scientifically rigorous approach.  The paper seems prescient 8 years later, and worthy of consideration moving forward.

Reading Group Summary: Alan Walks Wales

Readings:

Agenda:

  • Brief revisit of reading group and introductions of anyone new
    • Our time this week overlapped with faculty meetings
    • Core ideas for TotT; see previous week
    • Attendance: 7 people(1 professor, 4 graduate students, 2 undergraduate students, including 2 new faces)
  • Summarize papers
  • Discuss papers

Discussion:

We started with an overall discussion of the Alan Walks Wales project from miscellaneous outside readings by a few of the reading group members. This week, our two papers were very recent, so we only touched briefly on whether new tools or techniques (most specifically to dealing with open, large data sets) existed since the time of the paper’s publishing.PubsofBlacksburg-Poster-Maroon

Following the suggested definition of a wonky map from Alan Dix’s paper, we put forth as our own example the PubsOf poster shown here of our own local bars drawn by Brian McKelvey. It is “wonky”, as it doesn’t faithfully represent the town of Blacksburg but includes distortions that highlight key destination points (i.e., pubs).

With this example and the paper in mind, we talked about the definition of a “map.” One person suggested that a map depicted places that are roughly co-located with an emphasis on directionality, and another suggested a map was a graphical representation of physical places.

From the discussion of maps, a few interesting threads of conversation arose. We debated, especially for the purposes of travel or hiking, whether a map based on time-to-travel rather than physical distance might be possible. Traveling the distance up a mountainside takes far longer than traveling the same distance on flat ground. We also talked about examples like subway maps which commonly shrink the map of rails into more of a conceptual understanding of intersections, especially to fit on smaller pieces of paper or near the ceiling of a subway car.

We also discussed what it’s like to describe directions using landmarks to someone unfamiliar with the area, especially in the context of back routes or walking through alleys and grassy areas; sometimes we just say “follow me” to lead a stranger through campus instead of trying to describe such a thing. We also talked about on-the-ground assistance people give each other that traditional maps can’t provide, such as a trailhead marker that has a slider for its condition.

We talked about what it would be like if trail maps were read the same way as street maps, or if they were described solely in terms of landmarks. One of us suggested there are three ways to navigate by landmarks: solely on landmarks, landmarks plus direction, and purely on a sense of distance between landmarks. We also hit on the significance of general map literacy, going again back to A Walk in the Woods in which the author went into the experience with fairly robust map-reading skills, such as when he found a logging road to get around an impassable mountain in a storm.

We connected the two readings when talking about what maps could be generated from information collected on the ground. A map of the difficulty of the terrain could be generated based on the people actually traversing it. We talked about how user-generated maps might also capture those back routes that traditional maps miss that would normally spread by word of mouth only, or other issues like seasonal closures.

When discussing open data in classrooms, we hit upon the pros and cons of giving students entirely unexplored sets of data as open-ended puzzles to figure out. We talked about the need, also stressed in the paper, to clean the data beforehand. We felt a big issue could be the student not knowing how or where to get started with such an open problem. The data feeling irrelevant to the student could also be an issue. We talked about whether working with the data of hundreds of people in the set might make it more interesting to work with. However, that still might not make the data feel relevant to the student, and more isn’t always better. What the student cares about might not be something numerical data can capture.

We talked about two ways to approach data: either look for insights in an open-ended manner, or try to verify a hypothesis. The two are not always mutually exclusive. We talked about how readers of a blog might feel more connected to the story if they had access to tools to analyze and gain insights about the data contained in it, and specifically whether that would make readers feel more connected to the Alan Walks Wales story when they might not have before.

We talked about the potential of tying personal data sets, such as one generated by Fitbit, into publicly accessible data sets, such as meteorological data. Enriching data after the fact with additional public data would be an interesting area of study. However, we talked about the downside of highly localized data; going back to the meteorological data, weather can be very different in one area than it is only half a mile away, such as in a valley and on top of a foothill, and it’s less likely that publicly available data sets would match your exact experiences. Aggregating the data of many people with similar private data might crowdsource some of that specificity in an interesting way.