Bad Elf: delivering (very) affordable accuracy—simply—for mapping and asset management users.
The device is a Bluetooth peripheral GNSS receiver that can feed more-reliable and precise positions to any location-based app on your smartphone or tablet.
Having just returned from a visit to Bad Elf’s offices in Scottsdale, Arizona, I was itching to put this little device to the test. I had been hearing the name “Bad Elf” for years from environmental mappers, folks doing asset inventory for utilities, geohipsters, and surveyors who use it to hunt for survey control marks—they rave about it.
So, while waiting at the airport bus stop, I opened the box and fired it up. On my phone, I launched the free local transit system app using the phone’s built-in GPS/A-GPS. It showed my current “blue-dot” location as being in the middle of a busy highway, but once I connected the Bad Elf, the location jumped to right on top of the bus stop—dang!
More formal testing followed (with higher-precision options enabled), but the fundamental value of the Bad Elf was already evident. I could see why these little egg-sized pucks were all the rage with professionals and prosumers.
The Need for Enhanced Positions
Bad Elf has been producing “location enhancing” peripherals for about six years. The first products were produced to serve civil aviation customers. Most private pilots have taken advantage of a boom in aviation apps (produced by others) for phones and tablets; those provide active maps and charts and support simple navigation but rely on the stock positioning capabilities of the devices.
Phones and tablets often have both GPS and A-GPS: solutions that use cell tower data to improve time-to-fix (TTF) and to provide positions in environments that are not good for GPS (e.g. urban canyons, under tree cover, indoors). A-GPS is a wonderful enhancement for cellphones, and is also seen as a boost for 911 location capabilities.
While the position provided by a phone can sometimes be impressive, more often than not it will put your location as being meters out of whack, with a tendency to jump around. Your phone, using the combination of GPS and A-GPS capabilities, may suddenly decide it needs to recompute, and you will see wild jumps in location of tens of meters! This is not tenable if you plan to do serious mapping.
The first products from Bad Elf were plug-in GPS receivers (for 30-pin and Lightning connectors), focused on connecting to iOS devices, as those were predominant among aviation app users. Bad Elf has since become more “OS agnostic” with support of Android and Windows mobile devices.
Larry Fox, VP of marketing and business development at Bad Elf (and a private pilot), said, “Aviation apps are very popular for pilots; they can do many of things you previously had to do with paper charts, right there on their phone/tablet that has updated charts on-board. Then there is the safety aspect. With a glance, a pilot can see all of the airfields in the vicinity [of their aircraft] and quickly plot a course.”
Fox explained, “With apps, for aviation needs, 3-5 meters is mostly fine, not using it for glideslope, of course, as altitude is most challenging for GPS. But with the apps, you get everything on a nice screen and with a massive database of airfields.”
Bad Elf does not develop such aviation apps, but they provide what was missing: an independent GPS position that does not rely on cell data and does not jump around.
Fox noted that aviation is a limited market, and to move beyond it Bad Elf needed to add to their existing line of plug-in receivers to produce self-contained receivers that could operate standalone devices—and provide a range of higher precision levels. The new product line would also need to be Bluetooth-pairable with mobile phones and tablets, unleashing higher precisions to any app that uses location.
There are many apps out there for mapping, navigation, and asset inventory to help in hunting for survey control and more—apps where meter and sub-meter positions are perfectly fine. Interest in phone-based location was boosted by geocachers, people wishing to geotag photos, and social media apps. Marine was another market that implemented GPS location from even the early days of civilian GPS devices; there are many low-end, purpose-built marine receivers, even in the $500 range.
But mapping and asset inventory still often relied on dedicated hand-held data collectors with built-in GPS that could cost thousands of dollars, often with a premium on training and complex post-processing. Fox is a familiar figure to many professionals in the positioning and mapping community, having developed several successful solutions. Of his current work with Bad Elf, he said,
“The sweet spot Bad Elf inhabits is the marriage of the simplicity and power of mapping apps and precision at a great price point.”
There are three versions of the little egg-sized receiver that yield different levels of precision for different end-user needs. They can all operate as standalone devices, much like little backpacker recreational handhelds. They have backlit LCD screens that display location, speed, altitude, and GPS lock status; you can also do standalone or connected GPS data logging.
You can typically run them for about 16 hours before needing to recharge, or as much as 32 hours if the Bluetooth is turned off. They can log 100 hours or more of GPS position data. There is streaming NMEA position output to apps on your mobile devices. No subscription is required (except for external commercial correction sources, like an RTN, if used). The price point for each model goes up with added precision and features.
GPS Pro: This entry level GPS-only device can use WAAS (or other SBAS augmentation systems in different parts of the world like MSAS and EGNOS) to yield about 2.5m accuracy after achieving GPS lock in 45 seconds or less. Suited for general mapping and asset inventory, it costs $199.
GPS Pro+: This is dual constellation—66 channel GPS+Glonass—so it provides better performance in GPS-poor environments (canopy etc.). There is USB file access, and you can obtain altitude from both the GNSS and an electronic barometer (that you can calibrate). Accuracy is also around 2m, but up to elevations as much as 20,000m and at speeds of 1,600kph for the aviation market, and it can log as many as 200 hours of trip data. At $299, this model is well suited for mapping and asset inventory in mixed sky-view environments. Crews for such work can be affordably outfitted and can work in disconnected mode using whatever app you choose on nearly any mobile platform.
GNSS Surveyor: This has all the features of the other models, plus options for greatly improved precision. This model has an advanced UBLOX chip as its foundation, and it can accept real-time DGPS corrections from a local RTN or I.P.-enabled base station via the NTRIP client in the Bad Elf iOS app (more of that later) or any NTRIP client on the phone. The same Bluetooth connection that can feed corrections (presently limited to RTCM2.3) from the phone or tablet can stream back NMEA position data.
With a DGPS fix, this $599 device can provide reliable sub-meter (stationary, or 2.5 meter while moving) precision, enabling nearly any mapping and asset inventory projects that do not require centimeter-grade precisions. This model can also log and export static data for post-processing in the UBLOX native format or Rinex 2 or 3. Any post-processing package or online service that can process L1 Rinex will do. Fox recommends 15-minute static observations for post-processing, and some results have even come in sub-foot. As a surveyor, I did have to raise an eyebrow at the model name “Surveyor,” as we are notoriously a bit frosty about what constitutes “precision and accuracy.” But, as Fox notes, Bad Elf is making no claims that these are survey-grade instruments—these are mapping grade—and the word honors the term “surveyor” as the epitome of accuracy. (I joked with him about the word “doctor” in Dr. Pepper.)
Many surveyors I know do use small, handheld devices for certain needs, recon and estimating work, and to hunt for and inventory survey control and other site features. My tests were conducted with these applications in mind.
Of course, note that method and procedure are the key to achieving such precisions. Being aware of the nature of GNSS and how your immediate environment can affect results is very important, like multipath hazards, limited sky, and paying attention to quality indicators. As you need higher precisions, there is a premium on user knowledge and experience, and on this note Bad Elf has done an effective job providing excellent reference materials and primers on best practices, in some ways rivaling some of the high-end providers. You can see some of their resources here.
Speaking of names, I had to ask the obvious: “Where did the name “Bad Elf” come from?” As Brett Hackleman, Bad Elf co-founder and chief technology officer, tells the tale, the name is a result of a casual meeting in a pub in Germany between Hackleman and former IBM colleagues.
Brett is a private pilot and had been aware of the popularity of aviation apps. His previous work in software-development-enablement services (the connection of sensors and systems for industry segments such as automotive) gave him a strong foundation in the integration of GPS. He and his co-founders decided to pursue something along the lines of enabling aviation apps with improved GPS.
These go to eleven
In Germany for a conference, the founders were at a pub when the subject of a name came up. The phrase “turn it up to 11” entered the conversation, having been coined in the 1984 rock mockumentary, This Is Spinal Tap. In a famous scene, one of the rockers had his amplifier dial modified for 11 levels — so he could be louder. The team also noted that the German word for 11 is “elf”; now all they needed was a prefix. They wanted the name to reflect their goal of producing disruptive solutions, so they chose “Bad Elf.”
My test drive with a GNSS Surveyor yielded better results than I had expected. The application I had in mind was inventory of survey control marks; I used an iPad Mini and an Android phone to test.
On the iPad, I installed the free Bad Elf app and paired via Bluetooth. The app is not mandatory, but it has a lot of great features: settings, map, GNSS status, options to export static observations for post-processing, and you can set up an NTRIP connection to correction sources.
For both iOS and Android devices I did not have to do any configuration on the phone or tablet other than to set them up to accept “mock locations” (i.e. this allows the positions from the Bad Elf to supersede the onboard GPS of the phone and/or provide location capabilities otherwise not available on that model of iPad).
On the android phone, I installed a free NTRIP client (Lefebure) and set it up to receive corrections from a local base station that outputs RTCM2.3, and I configured the Bad Elf app on the iPad to do the same.
The positions from the Bad Elf (like many mapping-grade products) are in WGS84 (a functional equivalent to ITRF and IGS) and not NAD83, but users who work in NAD83 are (or should be) familiar with utilities to transform results. For my test drive I chose to see what the accuracy would be from observing both ends of an NGS calibration baseline and checking the inverse of the observed points (that I had good ITRF positions for) against published distances.
I first opened a free survey control app on my Android phone and used the Bad Elf to navigate to the marks. I was using only the WAAS capability at the time but landed within a meter of the marks. Then, connecting to the corrections source, I set the unit directly on the survey mark and watched as it gained a DGPS lock. I recorded a dozen or so shots in several sets (and again on a different day), plus several 15-minute and 1-hour static observations.
Results hit the published inverse mostly around two feet for the real-time results and close to a foot in post-processing. This was a clear sky situation, but a third mark checked was in moderate canopy, and results hovered around a meter. These are only a few sample observations, but I have talked with users who tell me they see similar results. The punchline is that with the phone GPS alone, the results varied by more than 10 meters.
For many GIS, mapping, inventory, and preliminary work functions, these units appear to deliver very affordable fit-for-purpose positioning.
Imagine putting something like this into the hands of students to teach them mapping fundamentals; to that end Bad Elf has worked with several education outreach programs. Imagine developing economies that need to map basic GIS thematic datasets. Imagine a huge utility finally able to afford to update very inaccurate (often scaled) legacy infrastructure GIS themes.
Everyone has a smart phone. Imagine what you and your crews could do if they “crank those smart phones up to 11.”