It costs a stupendous amount of money to send something from the surface of Earth to the surface of Mars, and there are severe limits on the volume and mass that you can send at any one time. In order to stuff the maximum amount of science into the minimum amount of space, NASA has had to get creative, with landers and rovers designed to be lightweight and foldable.
At NASA’s Jet Propulsion Laboratory, in Pasadena, Calif., engineers have long been trying to cram as much robot as possible into the absolute minimum amount of space, and a team of roboticists there recently showed us their latest creation: PUFFER, the Pop-Up Flat Folding Explorer Robot. It’s designed to pack down nearly flat for transport, and then re-expand on site to investigate all the places a bigger rover can’t quite reach.
The overall idea with PUFFER is that you’d pack a bunch of them along with the next Mars rover, and send them out whenever you want to go somewhere that it would be either risky or impossible for the larger rover to go. Maybe this is crawling along dunes of deep sand, taking a trip down the steep sides of a crater, or exploring little nooks and crannies where a larger rover simply can’t fit.
PUFFER’s small size and weight also open up some interesting possibilities when you think about sending more than one of them on a mission at once. Potentially, a lot more than one of them. Having access to a small swarm of PUFFERs means that you could set up robots to cooperate with each other, perhaps even to the extent of robots providing physical assistance to one another to do more comprehensive science.
One of the most exciting things about PUFFER is how it’s helping to bring some of the coolest robotics research we’ve seen over the last several years into the realm of practical applications. Most of the time, when we write about things like origami robots, the best that we can say is that in the abstract they might, at some point, be good for disaster relief or exploration or something like that. With PUFFER, JPL is taking the next step, saying, “Okay, how can we make these technologies actually do something useful in a real world environment, even if that real world is some world other than Earth?”
To learn more about PUFFER, we spoke with Jaakko Karras, PUFFER’s project manager at JPL.
IEEE Spectrum: Some of the research that PUFFER builds on, either directly or indirectly, includes origami robots, robots that fold themselves, and robots with tails. What other robots helped to inspire PUFFER’s design?
Jaakko Karras: I was a graduate student in the Biomimetic Millisystems Lab [at UC Berkeley], and the PUFFER concept draws a lot of inspiration from that research. PUFFER started when I came to JPL and started developing some early SCM [smart composite microstructures] prototypes of collapsible robots. I felt that there was a good use for the technology in the NASA context, particularly as an enabler for swarms of small planetary rovers. Much of the PUFFER folding chassis mechanical design was done in collaboration with Ron Fearing and his group, so there are a lot of their research themes rolled into PUFFER as well.
PUFFER’s ability to flatten itself is also similar to David Zarrouk’s robots, right?
I had a desk right next to David for a couple of years while I was at Berkeley, so I’m very familiar with his work on the STAR robot. David did a lot of very good research looking into mobility with sprawled wheels, and PUFFER certainly drew inspiration from his results as well. In particular, David has shown the ability to skitter beneath low overhead clearances in the sprawled stance, and we extend that to accessing areas beneath overhung rocks on Mars.
There have been a few other robots that take advantage of folding components, including things like wheels. Can PUFFER get even more foldable than it is now?
We’ve looked into the origami wheel research quite a bit, and we’re very interested in the concept for PUFFER! We’ve even done some early prototyping in that space, and hope to integrate some of those designs in the future. Having deployable wheels would definitely make PUFFER more foldable, both for increased compactness for integration on a spacecraft as well as improved mobility in tight spaces. Origami-inspired wheels would likely improve impact tolerance as well, since the wheels are typically the first point of contact during falls, and the origami wheels would provide excellent energy-absorbing compliance.
What kind of trade-offs are there in terms of reliability and durability when you make a robot foldable?
Foldable designs bring both benefits and challenges when it comes to reliability and durability. In the case of PUFFER, the folding chassis has a lot of flexibility and compliance, which allow us to survive impacts. We’ve driven PUFFER over 1-meter tall ledges, successfully surviving falls onto dirt and even concrete. We’ve also demonstrated survival when tumbling down a near-vertical 10-foot cliff. The group at Berkeley has demonstrated its legged origami robots surviving a drop from the top of a 10-story building.
One of the biggest challenges with these designs tends to be the durability of the flexure hinges, which need to be manufactured from thin, flexible materials. We often see tearing and fatigue-induced cracking at these locations. On PUFFER, we also pass copper traces over the flexure hinges in order to connect different parts of the chassis electrically, and design precautions had to be taken to ensure that these traces survived the number of fold-unfold cycles that we expect over a PUFFER’s lifetime. One design concept that we came up with for PUFFER was to laminate a really robust Nomex textile layer onto the rigid-flex PCB, and then use that as the flexure material, instead of the traditional polyimide material that’s typically used to link the rigid PCB sections. The Nomex textile is extremely compliant making it a great material for the folding flexures, and it’s very tear-resistant. We still pass our copper traces over polyimide film sections, but since these no longer need to provide the mechanical hinge function, they can be quite long for a gentler bend radius and long copper lifetime.
It sounds like PUFFER’s size is not finalized. Can you talk about the considerations that go into deciding whether to make the robots bigger, or smaller?
PUFFER’s size can be scaled either up or down, depending on the specific mission concept. Reasons to scale up might include wanting to accommodate larger instruments (or more instruments per platform), more batteries and solar panel area for greater power budget, or possibly for larger wheels and ground clearance for mobility in rougher terrain. Of course, a mission may want to scale PUFFER down as well. Reasons for this might include wanting to pack more units into a limited payload volume, for example on a very small lander. The good news is that the rigid-flex PCB chassis scales quite readily to accommodate a wide range of scenarios.
How would a swarm of PUFFER robots be more effective at doing science than multiple PUFFER robots operating independently?
There are a number of mission scenarios where cooperative investigation would be beneficial. One example would be a situation where you want to get into really adverse terrain, such as a steep cliff face, that is beyond the capabilities of a PUFFERs driving independently. Here, PUFFERs could instead link up in a manner similar to rock climbers working in tandem, and cooperatively climb the slope. Another example would be trying to access a deep cave network. Independent PUFFERs would quickly lose communication with the parent spacecraft. Working together, however, multiple PUFFERs could act as a repeater chain to relay data back from within the cave.
Can you describe a hypothetical mission scenario for a swarm of PUFFER robots?
The cave scenario above is a good example of a situation where PUFFERs could enhance a mission by providing access into an as of yet unexplored science-rich target. In this scenario, a larger “parent” rover would drive within say 50-meter of the cave mouth (keeping a safe distance itself). It would then eject multiple PUFFERs from an instrument-like PUFFER container, and these then unfold themselves. The PUFFERs would initially be guided to the cave through the parent rover’s cameras, but once they enter the cave, and the parent’s vision becomes inadequate, they would relay images from their own cameras back to the parent for processing. Most of the heavy autonomy computation would likely take place on the larger, more capable parent rover computer, with the parent sending commands down the PUFFER relay chain. Once the PUFFER at the front of the chain gets far enough into the cave, it would use instruments such as a microimager and small spectrometer to assess the inside of the cave for water history, habitability, etc.
What kinds of terrain are challenging for PUFFER, and how might the design or behavior of the robot be improved to handle such terrain?
Because of its small size, PUFFER often has to drive around obstacles that a larger rover would simply drive right over. Basically, PUFFER’s lower ground clearance means that the robots need to be a little more creative with path planning when driving in very uneven terrain. On the other hand, this ability to fit in between obstacles like rocks is what allows PUFFER to get into all the tight nooks and crannies that a larger rover can’t access, so there are trade offs. Having expandable origami wheels could provide the best of both worlds: the wheels could expand to traverse larger obstacles more directly, and collapse to access the confined spaces.
Terrain-specific wheel and tail design can contribute tremendously to PUFFER’s mobility. Current prototypes are being designed for the types of terrains found on Mars, and the team is optimizing its wheel designs for these types of surfaces. Recently, the team added microspine features (inspired by work done with spines for various climbing robots) to PUFFER’s wheels, and saw a remarkable improvement in incline performance on its Mars analog test terrain when compared to results achieved with more universal wheel designs. Since PUFFERs would be tailored to specific missions, it makes sense to optimize the mobility features for the target terrain.
What are you working on next?
Now that we have good PUFFER prototype hardware, we’re starting to look into the autonomy software that will enable a parent spacecraft to dispatch and coordinate swarms of them. We’ve only just started on the autonomy, but it’s clear that there are many exciting challenges to address there, particularly in the multi-PUFFER cooperative behaviors that we hope to develop.
[ PUFFER ]