A little more than a year ago, Swift Navigation started a Kickstarter campaign for a $995 RTK GNSS receiver named Piksi. The goal was to raise $14,000 for the project. By the time the Kickstarter campaign was ended, Swift raised $166,097 from 303 backers.
I wrote an article about the Swift’s Kickstarter campaign in September 2013, a few days before the fundraising period ended. Following is a two-minute Kickstarter promotional video describing Piksi.
This week, I took the opportunity to catch up with Swift to see how the project is going and where the company is headed.
I spoke with Tim Harris, CEO of Swift, and Fergus Noble, chief technology officer. Tim said they delivered most systems to their backers in April (2014) as well as issuing “very few” refunds, which is not unusual for Kickstarter projects. As of today, Tim said there are about 1,000 units in the field among 350+ users. He said they still consider the Piksi to be in beta testing phase and expect production units to start shipping next year.
“We are in RTK reliability mode now” — Fergus Noble
Fergus said that the hardware design is stable and hasn’t changed since they shipped Piksi last spring. He said the focus has been on refining the RTK firmware to make it robust and reliable for professional use. He admits that “covering all corners” has been a challenge and, according to Swift’s blog, “The RTK software has been a tougher nut to crack than we originally anticipated.” This is especially ambitious, since Swift said they developed the RTK firmware based on published academic technical papers, combined with their own techniques (preparing patent applications) to run on a low-power hardware platform. Basically, they started from scratch, and without any RTK veterans on their team. You gotta love their guts.
Apparently, the venture capital world liked what they saw. A few months ago, Swift, based in San Francisco, secured $2.6m in seed funding from Fall Line Capital, Felicis Ventures, Kal Vepuri, Lemnos Labs, Qualcomm Ventures and VegasTechFund. With the additional funding, Swift has increased its headcount from 4 to 9.5.
The Devil Is in the Details
Start-ups like these are a cool story, but at some point the rubber has to hit the road. Piksi is a long way from being an RTK product that you and I use for GIS and surveying, and Tim/Fergus are humble enough to admit this. The founders see the Unmanned Aerial Systems (UAS) market as their key customer base due to Piksi’s micro-size and low cost. They say this is reflected in their current customer base, which is comprised largely of UAS users, then surveying and agriculture, with the remaining 20 percent being a variety of apps like augmented reality. I agree that UAS is the right segment for them to focus on. At this point, the limitations for general surveying and GIS use are too great:
- Must use a Piksi RTK base due to the proprietary protocol. In other words, not compatible with existing RTK base stations or networks.
- Baseline length (distance between the RTK base and rover) must be very short. They haven’t done much testing even with a one-mile baseline.
- RTK initialization is spec’d at 15 minutes, although Fergus said it’s typically 7-10 minutes.
Since Piksi is only a single-frequency receiver (L1), it has the same limitations as other L1 RTK systems we’ve seen marketed over the past few years — short baselines and long initialization times compared to dual-frequency receivers. While these limitations are significant for surveying and GIS users, they aren’t as significant for UAS users, since UAS users require line of sight to the aircraft at all times and UAS are operated in a clear-sky environment.
If you’ve used RTK, you are aware of the importance of data communications between the base and rover. As part of its kit, Swift supplies license-free, 915MHz frequency-hopping radios to facilitate communications between the base and rover for testing. However, the founderse make it clear that they aren’t in the data radio business. They expect that most Piksi users will incorporate their own communications technology. Swift is discussing incorporating GSM (mobile phone) modem capabilities in the future.
In 2015, Swift plans to roll out the Piksi in production quantities, so Fergus said their focus is on creating a robust and professional RTK system. Of course, the burning question is if/when Swift will incorporate a second GPS frequency into its receiver to take advantage of the faster RTK initialization and longer baseline distances that a dual-frequency receiver offers. Swift isn’t willing to talk about that right now. Said Tim, “We’re holding our plans close to our chest.”
Thanks, and see you next month.