News

SubPos Kickstarter for Sub-Meter Indoor Positioning System

August 02, 2016 by Jeremy Lee

The SubPos Ranger is a Kickstarter campaign to manufacture an open source Indoor Positioning System intended for hobbyist robotics applications that need sub-meter positioning accuracy.

Let us introduce you to Blair Wyatt, the engineer behind the SubPos Ranger: an open source Indoor Positioning System platform.

The SubPos Ranger is a Kickstarter campaign to manufacture an open source Indoor Positioning System for applications that need sub-meter positioning accuracy. The rangers use the Atmel AT86RF233 for phase-based distance measurement using radio signals, which are accurate to within 10cm in ideal conditions or 30cm when obstructed.

The primary aim is to create an open source platform for small robotics platforms. A complete 3D positioning setup includes four ranger boards and one ranger/expansion client board. If their campaign is successful, they estimate delivery in October.

Blair Wyatt is the founder of SubPos, the Australian company behind the SubPos Ranger. A simulation systems engineer "by day" and born tinkerer at any hour, Wyatt founded SubPos with his wife, Sue Moon, just this year.

AAC's Jeremy Lee had the opportunity to ask him some questions about the campaign.

Is this your first Kickstarter?

BW: This is our first Kickstarter. It has definitely been a great lesson in the marketing and planning side of what it takes to create and release a product, which is something we still have a lot to learn about. It was something that was unfamiliar and scary, to which we thought, why don't we jump head first into the deep end.

How's it going?

BW: We have a system that works well and we want to get it out there for people to use, which is certainly the toughest part of the whole development process. With the Kickstarter missing the target, we still want to create a small-scale production run for people who are interested and see where it goes from there. This will be used to further develop useful applications of the technology through the community—we can then hopefully move onto a more marketable and revised product at a later stage.

So you're an Australian company. Are you doing the production in Australia? If so, is the local infrastructure sufficient? 

BW: Under the Kickstarter run, we did have plans to completely produce in Australia, except for component procurement. We will still push for Australian production for a smaller-scale run, but this will depend on negotiations, demand, and budget.

Australia does have sufficient manufacturing capabilities; it is, however, limited by price. To make local manufacture worthwhile, you need to be prepared to increase the size of production to a point where you can take advantage of economies of scale and be happy with smaller profits to release it at the price point you set. Doing things locally, however, does allow you to control the whole process, yourself—which reduces the complexity and risk.

Ranger with client expansion board and Arduino Teensy.

Let's talk about the SubPos Ranger's specifics. Measuring distance with phase is very sensitive to multipath interference and path fading. Has the dual antenna design been enough to completely eliminate that? 

BW: The system is really designed as a hobbyist platform for the development of applications rather than a full-scale commercial setup. So the system, like many others, is definitely affected by multipath effects when there are large and dense objects in the way, so operation through walls is still problematic. However, these issues can be overcome by having more devices situated in different rooms to provide greater coverage. The dual antenna design helps overcome the issues associated with smaller (possibly moving) objects and reflection effects.

Is there a particular semiconductor device that made SubPos possible? 

BW: The Atmel AT86RF233 has an onboard Phase Measurement Unit which is used to perform the raw distance measurements (through phase difference measurement from the transmitter to the receiver).

There is also a Time of Flight module that, while isn't accurate enough to obtain sub-meter accuracy, does allow you to compensate for full phase turnover at certain distances. So using these two in unison allows you to obtain accurate and precise measurements.

The Ranger client locates itself in 3D space by measuring the distance to multiple Ranger Boards using radio waves, similar in concept to the GPS system but at much closer range.

How hard would it be to integrate the SubPos into an existing robot or prototype device? Does it speak any standard protocols? 

BW: Once you place the client on the top of your device so it has the clearest view of the nodes, the system then spits out a raw positioning measurement via a hardware UART or USB connection. This measurement can be output as an NMEA string if using geo-referenced measurements or just an offset x-y value. The measurement value can then be input into various autopilot systems for easy navigation.

Being open source, you can also change the way this information is sent out from the device. So the plan is really to provide the hardware sub-system and then support a community which will further develop other interesting applications.

Is there anything you'd like the maker community to know about SubPos?

BW: Apart from distance measurements, the SubPos Ranger is designed as a multi-purpose radio frequency development board for many different applications. Some of the examples we have shown allow you to perform passive motion sensing, or use it for general purpose communications, but there is certainly something out there that hasn't been thought of. So the message we are trying to send with the device is that, as makers and engineers, we shouldn't be limited to what is presented to us. Continue to explore and experiment with different ideas and don't be afraid to try something new.

Visit the campaign's page to learn more about the project. It's entering its final hours at this time!

1 Comment
  • S
    Stephen Young August 19, 2016

    Aaarrrgh!  Why does this article coming out 2 weeks AFTER the campaign has ended.  UNSUCCESSFULLY.  I would have contributed, but obviously, they lacked in marketing.  A great device, hopefully they will try again.

    Like. Reply