I spent the last week in Kabul, Afghanistan meeting up with our local partners, clients and making new friends and associates in the geospatial community working on various projects and initiatives in Afghanistan.  It was truly a great week.  One of the highlights of the trip was the chance to get out of the confines of the hotel and do some field data collection and product testing for fulcrum.

I had set up a couple of forms and test projects before leaving Florida, and our office team prepared our 1:5000 scale dataset of Kabul in MBTiles using TileMill so that I could download via fulcrum an offline map for use in the field while testing.  I didn’t really have a well-thought out plan on what to test, how and when, I just wanted to get out and use it and hopefully uncover real use-case issues that our engineers and designers could use as feedback to continue to improve the product.

Below is the results of the week – approximately 465 POIs collected in the field in just 6 days.  Mind you, I wasn’t out of the office for a full day of working, but nevertheless, the productivity and capabilities are encouraging.

I was most impressed with the spot-on accuracy of the iPhone and iPad on-board GPS.  I tested data collection with only WiFi (in the hotel courtyard), cellular only (Roshan, Etisalat), both WiFi+Cellular, and finally airplane mode.  The accuracy relative to our offline map was outstanding in every configuration.  I further tested syncing via cellular and wifi, and to my surprise, the cellular network outperformed the WiFi syncing.  The hotel WiFi was only 802.11g and unless I was directly within line-of-sight, speed fluctuated widely.

To be fair, many of the POIs I collected were not heavy with data attributes and most did not contain pictures.  I did collect enough pictures to get a grip on the relative difficulties in uploading 10+MB at a time of pictures in an environment where the internet infrastructure is primitive or non-existent.  But, this was exactly the point of the testing – real-world, real use-cases and thus the offline mapping capabilities of fulcrum performed exceptionally well.  Collecting data for days or weeks in offline mode is exactly how fulcrum was engineered to perform.

I’m sure our engineering team will be able to get much more diagnostic data when I get back to the office and they have a look at any logs or metadata captures during the testing, but it is clear to me more than ever that while we have alot on the roadmap for fulcrum, it works as intended today in the environments for which is was designed.