The next generation of robots with Paul Ekas

Paul Ekas is president of SAKE Robotics, who's vision is to deliver the lowest cost, high dexterity robotic grippers that enable next generation robots.

author avatar

15 Dec, 2019

Paul Ekas

Paul Ekas

Paul Ekas is president of SAKE Robotics, who's vision is to deliver the lowest cost, high dexterity robotic grippers (“hands”) that enable next generation robots. In this podcast, Paul talks about his EZGripper and how he designed it to be low cost, lightweight, robust and to offer reliable gripping of small and large objects.

The EZGripper by SAKE Robotics is a tendon based gripper using Dyneema tendons and aluminum oxide eyelets to make it durable and able to handle rough environments.

The EZGripper is under-actuated, the fingers stay straight when picking up small objects and wrap around large objects. You can control position and torque allowing you to grip soft or hard objects and to do so gently or firmly.

The next generation of robots w/Paul Ekas


Per Sjöborg, host of the Robots In Depth podcast, interviews Paul and talks about his EZGripper. Below is a transcript of the interview.



Per: Welcome to the podcast version of Robots in Depth.  I'm here with Paul Ekas from Sake Robotics. We’re going to talk about robotics in general but also specifically about the robotics hand he's developing so welcome to the show. Could you start with a short description of your hand here?

Sake Robotics EZGripper

Paul: Thank you Per. I've been developing the Sake Robotics EZGripper for the last couple of years, gone through a variety of iterations. This has some unique capabilities. It is very easy to operate. You put it on to an object and you tell it to close and it will hold that object and not drop it. It can be in an unstructured environment so it doesn't matter what kind of object it is. It is also designed to be super low cost so when 20 years out the intent is to have it be the lowest cost architecture for a very long time and a platform to build on. That is the Sake EZGripper and it has been introduced in the last few days.

Per: We're here at ICRA 2015 in Seattle and we'd like to thank the IEEE and the organizers for providing us this space and all these systems they have given us. You talked about low cost. You talked about architecture. Could you give me the key points that you focused on to bring about this low cost and making it a platform?

Low cost robotics platform

Paul: When I looked at grippers in the market my desire is to have grippers do something like dishes. I don't want a robot friend. I want a useful robot to get stuff done. I wanted a robot that could stick the hand into the dirty water and do that every day at a very low cost. To achieve that I looked at the types of hands that could be built and about half of them are tendon based, half have the motors integrated into the joints. Having motors integrated in the joints requires a lot of complexity and miniaturization, has reliability issues and if you plunge it into water it has some risks. I like the tendon based ones. The human arm is tendon based so I pursued that. A challenge in tendon based robotics is where there's a variety of patent-pending technologies in this but one of the core patent-pending technologies that differentiates this from all other tendon based hands and grippers is I'm running Dyneema which is common but I'm running them through super hard ceramic eyelets. I'm using aluminium oxide here which is the same material as the sapphire face of your watch. That allows to have a low cost architecture that is very high reliability and many cycles. That’s an aspect of it.

“When I looked at grippers in the market my desire is to have grippers do something like dishes”

Second aspect is I am using plastic for building this. This is 3D printed with SLS but it's all been designed for injection molding so the fundamental cost structure of this in volume production is very low. It’s been designed for easy assembly, easy manufacturing, easy repair and just dropping the cost structure both from the purchase price and the maintenance price.

Per: Because if you open up with a lower price you really reach new markets that doesn't have the ability to pay what you have to pay today for a gripper which can be a lot of money.

Paul: It can be a lot of money. If you’re in the robotics a lot of the grippers out there got lots of fingers, lots of dexterity, really research oriented. This is a design that you put it on an object. It conforms to the object. Picks it up without hassle because the challenge in gripping with robots is getting the grip to the object. At the ICRA they have the Amazon Challenge. The best teams in the world getting the gripper to the object is a major challenge. When it's there it ought to pick it up, grip it, not let go until you want it to let go. It should be easy to use and that's what I've designed into this permutation of grippers.

Per: Also if it's also built to be serviceable because that's of course also very important that you can actually maintain it. You talked about it being suitable for a non-research environment. You’re talking about doing the dishes and having it in dirty water all the time. Could you talk a bit about the considerations you had to do to make it operate in those kind of conditions?

Non-research and other environment robotics applications

Paul: The architecture has received interest from space where it's incredibly lightweight to undersea where its materials all work very well for. It works across there depending on the type of motor put in. This is a winch based motor on the system so you can really put any kind of motor in there to deal with that. Dealing with dirt you want it to be able to drain, not have any things in it that really are dirt sensitive. It does have bearings in the joint. They’re sealed bearings and if you're in different environment like space, you don't want the lubricants of the bearings so you go to ceramic bearings. All the sizes associated with this are readily available. The same thing with undersea. You can literally put it in chemicals, rinse it off. I can take the motor off which is designed to be removable easily. I could put it in the dishwasher and wash it.

Per: That also open ups food preparation. It opens up medical fields. Have you looked into any opportunities in making this sterilizable for the medical field?

Paul: I've researched some of it. They use some amazingly wicked chemicals on there and the fortunate thing is the technologies I'm using in here are all capable of dealing with those types of chemicals. The weakest thing I'm using from a chemical perspective is nylon. This can be fabricated out of any material. I've actually submitted it for fabrication in titanium who is a thousand times more expensive.

Per: But if that's your application it is probably cheap anyway.

Paul: If that's needed it can be done but the Dyneema actually is incredibly chemical resistant. You can stick that in nail polish remover and it won't dissolve. The bearings if they're made out of ceramic don't care. All the metal is stainless steel in it so it really is fundamentally chemically resistant design except for the mold of the hand itself which can go in any direction based on the customer's needs.

Per: It can both be cheap when it needs to be and it can also be durable depending on the chemical in the environment but still to be reasonably low priced. Have you done any gripping test? This is a gripper. It’s got to grip things reliably. It’s got to perform this task. Can you tell us a bit about metrics around this?

Robotic gripping test metrics

Paul: Gripping metrics is an interesting space. At the show NIST is starting a major project which is actually around grip strength. When I demoed to them the capabilities that I had they wanted me to participate unfortunately I'm still in the early stage from that but it's going to be a long-term thing. How do you measure grip because what people found at the show very exciting is about I can grab an object that is multiple kilograms wide. I can lift that, maintain it but at the same time it can go right around your arm, grip it with the same strength. I've been actually gripping a banana and not damaging it. The style of grip that I'm able to achieve with this is actually really a soft, firm grip. How do you characterize that? People have been using parallel plate grippers and they say how much force can it put on? It has to put on a lot of force because it's not necessarily using good technique to grasp something. Measuring grasp capability is an interesting challenge and the government's actually funding something to deal with that so I look forward to working with them and characterizing what I have with that context.

Per: If you hold it up a little bit higher we see this and it's about the size of a human hand is it also scalable? Could you do it in different sizes?

Scalability of the robotics technology

Paul: The pending patents have the point of view on this that it is really scalable to any sized structure that's using links and joints whether it's a quadraped or a humanoid or for doing the legs in the medical space might be for actuation around ankles. It's a very scalable technology from that perspective and has a lot of value in there. I focused this on the hands because the challenge that you have in hands that you don't as you scale up is space. You don't have a lot of space to work with. Putting little tiny motors in the joints isn't very easy. As you get bigger it's easier to deploy potentially other technologies that are more suitable than just moving motors to joints in many cases.

Per: You would use this in this size and down rather than this size and up.

Paul: There is an application for this in a larger size and I like to refer to this as a vision that I have of called the monkey robot which is a four-limbed robot but all the motors are pulled to the core. Why is that good? Because it makes your inertia centered in the robot. Let’s say that you have a robot that's servicing satellites and it crawls on the outside of a satellite. You don't want have a big weight moving around when it's moving its arms. It also allows it to move arms more easily without the inertia and it can hold its arm out with less supporting weight. That’s a challenge in robot arms because this is hard. They have weight limits of what they can carry because they're carrying a lot of weight in their motors. There is value in having it in the right types of systems. It’s not everywhere but it has some real advantage as potential when we get the smarts to build monkey robots that do useful work in unstructured environments.

Per: Is this available now?

Paul: It is available now. I just released this to the public three days ago. I had the great opportunity to be at ICRA in Seattle with Robotis. I reached out to them. I'm using their motor in the gripper. The motor is a Dynamixel MX64 which has lots of power. I underutilized the power for reliability. I reached out to Robotis and explained what I had. They invited me to the show and have taken great care of me. It’s been a huge opportunity for me. Then at ICRA we have the top tier of Robotis who come and give feedback and I've had a very exciting reception. It’s been great for me.

Per: Any comments you've had from the people attending and that you'd like to share?

People’s reactions and comments

Paul: The people's reactions are very positive which is great. They like the simplicity. It’s a very simple design which achieves the objectives that I want. They’re very surprised. They see this thing. It looks like a gripper out there and then they start picking stuff up with it. They can pick up a pen on the table or they can pick up a liter bottle of water and shake it. It conforms to different shapes and sizes. I think when people start interacting with it they enjoy it. I often have people grip their arm and they can feel the grip. They are very impressed by that.

One of the funniest ones I had is I have the gripper on a battery system so I can easily grip things. People can hold it and try it. Two guys walked up. I showed him this thing, handed it to him. He turned to his friend, stuck it right on his friend's face and grabbed his face. He didn't ask if it would be safe or anything. It is safe. It’s human safe but that was pretty funny.

Per: You should probably check with the designer before you grip somebody's face. If it fails that would be horrible.

Paul: We are not recommending that but it was fun to watch.

Per: Could you put this in context? I mean you're operating in the robotic space. Which kind of robot do you see this being attached to and what will that robot do for someone?

Use cases of the robotic gripper

Paul: I've got a variety of visions of robots out there and we all do because we've seen Hollywood. I’ve had a great friend of mine died of ALS, a brilliant guy. He’s in his wheelchair towards the end and if he had a robot gripper on the end of a robot arm that could be doing useful work for him which I think we can pull that off. It needs to go out and grab whatever is being desired and bring it up and work with the person in a safe environment. I see on the medical side that this can actually really add some great value and I love that. It feels good and that's fun. It’s great to talk about. I see at the extreme end of business potential is stuff like the Amazon challenge. They have an incredibly structured environment. They have all the data. They got the compute power and they have a lot of pick in place and this is a perfect fit for that. Get the gripper to the target not a hard problem and when it grips, it holds. It doesn't let it go and it puts it where it's supposed to go. It lasts for lots of hours and it's cheap to operate. That’s a significant business for robotic arms.

Then the vision that is I'm a lazy guy. I never worked so hard to automate everything but doing dishes, folding laundry. That is just useful stuff to automate. It is very hard to do and as robots get better at working in unstructured environments doing unstructured tasks that'll be an amazing set of products that can be brought into human society. People will buy them like crazy and never think twice after they've got them how wonderful they are. Someday.

Per: You're saying that you want to do this low-cost. Could you tell us anything about the cost right now and how you see the cost developing over the coming couple like three to five years?

Cost development in a few years

Paul: I'll be a little careful about that but the design intent from the very beginning is, my background is semiconductor industry. Relentless cost reduction and value increasing. When I looked at the robotic manipulator space I said, if I'm starting a company to do robotic hands, manipulators, in 20 years how do I have a cost advantage? It’s not that it will take 20 years to get there. I am right now by far the most capable at the lowest price of a robotic gripper. I can see that architecture maintaining its cost competitiveness for 20 years. How do you do that? You make sure that the materials you use are easily sourced, that the assembly of those materials is easy to assemble and the cost structure is of all the different parts of it is low cost. That is how you make a low cost product you design for low cost.

Per: Could you mention a price that if it cost that 10 years from now or 20 years from now you'd be really proud? That’s kind of the high end goal you have when it comes to costs?

Paul: I'm going to do a couple of things for you. I'm going to put one direction and then another direction. What I've got right now is the electronics are all in the motor and there's not a lot of sensor technology out on the hands. That sensor technology is going to evolve and there's tremendous opportunity there. That's going to go down in price dramatically. That’s a huge evolution that can occur along there and I've got a great background to be driving that at a corporate level. The rest of this is it's a box of Legos. I've got Dyneema high-strength line in there but that's essentially a fishing line. It’s already a commodity. I've got the ceramic technology in there. I won't go into detail on that but it's not an expensive thing. The plastic injection molding is a box of Lego's and it's got bearings in the joints. It’s got bolts that assemble it. It’s been designed for easy assembly so I won't give you a cost target but it's a box of Lego's for a robotic gripper that can do amazing things. Like I said in 10 years can I maintain the price? If the market goes down in the pricing then it's enabled by the technology that I've got. The goal is that the hands are ubiquitous, that they are everywhere and that there's a market that allows the price to go down and maintain a very strong business. This technology can ride that wave.

Per: Have you mounted it on any robot and if there's a researcher out there that's saying would this fit my robot? What kind of sizes are we talking about that you would be comfortable putting it on?

Adaptability on other robotic systems

Paul:  that’s a variety of questions so I'll answer different pieces of that. The first robot that I've mounted on is from Robotis. They’ve reached out. I said I need a robot arm and it would be ICRA. They just sent me a robot arm. Just huge opportunity.

Per: They are a great partner.

Paul: I’m using the Robotis Dynamixel MX64 and the beauty of this product in the robotics market is almost everybody in the research community who works with robots knows how to program these. They’re easy to do. It’s very adaptable into other robot systems. The only thing that really needs to be brought together to hook and other robots is a flange that goes from this servo to those other robots. I'll have an open-source reference design for the mount that goes this direction and they can 3D print something that goes to any robot. The second thing is voltage conversion. I have voltage conversion circuits that go around this and then if you can use the Robotis control for this which is RS-485 which highly reliable, standardized across a variety of different systems. It’s easy to integrate in there. Early on what I'm going to be doing is engaging customers around their robots and just building the interfaces to that. It’s not difficult to do and it's just something that will be driven by customer demand.

Per: The size of the robot we're talking about what is the smallest robot you'd be comfortable putting it on right now?

Paul: Let's look at a robot arm. A robot arm at some distance it can carry some weight. There's the big evolution of robot arms from huge industrial robots to now we're getting to these very small robot arms that are safe around humans etc. They have limited capacity for lifting stuff. Call that anywhere from two kilograms to six kilograms. That’s a reasonable range to have the discussion.

Per: We haven't mentioned the weight of it.

Paul: A key point is when the robot gripper is on one of these robots it's consuming that capacity. If you have a two kilogram gripper on a robot that has a two kilogram capacity you can do zero work. This is 350 grams. It’s ultra-light. That allows it to go on a range of robot arms that are smaller than most competitive products out there.

Per: And still have a useful payload that is significant.

Paul: Then so the question is as you go up to bigger and bigger robots how long is that still valuable? The biggest robots I'm seeing at this show are that are working in these kind of environments are things like the Baxter and UR5, UR10 etc. which are very capable robots. This is more capable of lifting larger objects than their standard grippers that they come with which are generally little pinch grippers. It can integrate nicely into those bigger arms and have plenty of payload capacity for them.

Per: Anything you'd like to close with? Any points we haven't mentioned for somebody interested in the hand and utilizing it?

Paul: Number one is we put up the website a couple of days ago SakeRobotics.com. It’s a developing a website, developing product. We can deliver a product in about a month to customers. Looking for early customers to engage. You heard all the value propositions in there and I'm excited about going forward and meeting customers, understanding their applications and being responsive to their needs in developing this technology forward into the robotics market.

Per: Perfect and thinking I love anything low cost. I love anything that's kind of modular and getting more robotics into the hands of more people to be able to experiment and develop things is very interesting. Thank you for taking the time to do an interview and I'll hope to hear lots of interesting applications from you in the future.

Paul: Thank you Per, this has been very good for me and I appreciate your opportunity to do this.

Per: I hope you liked this episode of the podcast version of Robots in Depth. This episode is produced together with Wevolver. Wevolver is a platform and community providing engineers informative content to help them innovate. It is how engineers stay cutting edge. Aptomica is the founding sponsor for Robots in Depth. Aptomica runs anything in modular robotics. Dream, rent, build. Visit Aptomica.com to connect. I am your host Per Sjöbor. Until the next episode thank you for listening.

END OF TRANSCRIPT


Follow the podcast on Apple, and Spotify.

The Wevolver Robots in Depth podcast is published multiple times per week. If you, or your company is interested in supporting this podcast and reach an audience of professional engineers and engineering students, please contact us at richard [at] wevolver.com 

 

15 Dec, 2019

I interview CEOs, founders and CTOs from successful and up and coming robotics companies, as well as researchers, investors and thought leaders within robotics and AI. You'll learn about the latest technologies that are being developed at innovative robotics companies, the lessons learned by incredi... learn more

Stay Informed, Be Inspired

Wevolver’s free newsletter delivers the highlights of our award winning articles weekly to your inbox.