Over the past few years, you’ve likely heard mobile manipulation is a trend to watch. Perhaps advances being made in a variety of enabling technologies – sensors, controls, power – are finally making mobile manipulation a reality. There have been at least three mobile manipulators introduced within the past month. Boston Dynamics certainly had the most interesting take with its Stretch robot, while Desmasa launched a more traditional mobile manipulator with its MCR+.
The third on that list is a new collaborative mobile manipulator from RBR50 company Waypoint Robotics and Productive Robotics, a Calif-based developer of cobot arms. The system integrates Productive’s OB7 cobots onto Waypoint’s Vector omnidirectional autonomous mobile robot. I visited Waypoint about a week ago to get a sneak peek at the system. The prototype I saw featured a Vector 3DHD and the basic OB7 cobot, but the mobile manipulator can also integrate Productive’s other 7 degree-of-freedom cobot arms – OB7-Stretch, OB7-Max 8, or OB7-Max 12 – depending on the application. The system uses Productive’s standard teach pendant.
“There’s a difference between us saying we have a mobile manipulator as a standard product compared to we have an option to have a mobile manipulator,” said Waypoint Robotics CEO and co-founder Jason Walker. “The mobile manipulator [with Productive Robotics] is something we’re integrating a lot more extensively and tightly.”
Walker and Productive Robotics president Zac Bogart first met at ATX West 2019. Both companies will be selling the mobile manipulator. Walker said this will benefit customers and integrators who are accustomed to working in particular environments.
“If they already know the language of the OB7, they can add a coding block to the teach pendant that tells the Vector to drive to a certain location and do whatever it’s supposed to do,” said Walker. “It’s a control paradigm that lives in the OB7. People who are familiar with that ecosystem don’t have to go learn a whole new ecosystem. It can also work the other way around, too, for those who are familiar with Waypoint and the Dispatcher software.”
ROS integration key
Walker talked at length about how Productive’s 7-DoF arms and Vector’s omnidirectional capabilities will enable the mobile manipulator to work in tight spaces and make unique movements. But there are plenty of other 7-DoF cobot arms. So why did Walker choose to partner with Productive?
“We’ve integrated a bunch of different arms, but the OB7 is energy efficient, it’s ROS-native, it’s made in the USA. [Productive] is a scrappy startup about the same age as us,” said Walker. And it’s a company capable of understanding the opportunity and the unique benefits of mobile manipulation.”
The mobile manipulator doesn’t rely heavily on ROS at the moment, but Walker said it offers a great path forward for even tighter integration. Within ROS, the kinematics for the entire, combined solution can be defined and leveraged to help coordinate motion between the manipulator and the mobile base.
“We can combine setup tasks for each device into one task. That will lower the effort required for somebody to set it up or reconfigure it or move it around from one location to another to do a different job,” he said. “The system also has two computers – one for the arm and one for the mobile robot. The more we can get the two things merging closer together, we can remove one of the computers to further reduce cost.”
While demoing the system, Walker set up separate waypoints for Vector in its software and the OB7 by using its onboard buttons and moving it through space. However, Walker said there’s no reason why one button can’t make both waypoints, highlighting another way ROS can help take interoperability to the next level.
Power consumption is key for mobile manipulators
Walker said one of the biggest problems with mobile manipulators is power and energy consumption. He said nearly all the mobile manipulators he’s built as one-offs in the past required a spare battery and additional integrated battery kit.
Walker said Waypoint approached the power problem from a number of angles. One of them is by using Waypoint’s EnZone wireless charging system. The mobile manipulator is capable of supporting two EnZone receivers, meaning double the energy can be delivered. He also said the OB7 operates natively at 24 volts, which is the same as Vector. Walker said his mobile manipulator can get at least six hours of run time.
“We’re committed to, over time, cost reducing this thing,” said Walker, who isn’t ready to share the price tag. “A big part of that is taking out all of the extra power conversion stuff, and doing a version of it that’s natively 24 volts. Instead of going through converters, we’ll feed battery power straight into the arm. And that’s going to significantly reduce the cost and size.”
Walker claimed Productive’s cobot arms are also more energy-efficient than the competition. “When we do a UR5 integration, we usually use a 1000 watt inverter. We leave the AC-to-DC converter that’s built into the box and then add an inverter so that you can use either DC power from the robot or AC power from the wall. [In this new mobile manipulator], we’re doing it with 400 watt inverter. The fact that it can be done with that small amount of power is huge. We’re lucky to get 4 hours out of UR5 mobile manipulator, so it’s just more efficient.”
Mobile manipulation has reached an inflection point
Workers can use this mobile manipulator for a variety of tasks, including machine tending, quality assurance sampling, material replenishment, and many others. We recently saw Boston Dynamics’ Spot quadruped grab and drag a concrete block. Walker said they’re currently working on similar functionality.
“An immediate use would be the ability to drag a tote off of something and onto the robot,” he said. “One of our targets are companies that don’t want to revise their environment, or want to as little as possible. If this robot can drive up to a counter and pull a tote off onto it, that’s less work than setting up a conveyor or a pickup and drop off station.”
Machine tending, parts kitting and bin picking are three examples of applications where a mobile manipulator can be deployed to complete the task. I’ve seen many mobile manipulators at trade shows that combine a mobile base from Mobile Industrial Robots (MiR) and a cobot arm from Universal Robots. Of course, both companies are owned by Teradyne. I recently asked Greg Smith, president of Teradyne’s industrial automation group, if we’d see an official mobile manipulator from the group in the near future.
“There’s an interesting economic challenge around arms on mobile bases. When people look at doing automation, they look at it from an ROI perspective. How long is payback going to take? How much utilization does the automation get? If you take a cobot arm and stick it on top of a mobile robot, then you’re in a situation where when you’re using the arm, the base isn’t adding any value. And when you’re using the base, the arm isn’t adding any value.”
“It has to be a much higher value application to justify the spend because you’re only using half the robot at any moment. There’s some challenges in terms of getting the right price point to be able to do that. But we also see some interesting niches where even with today’s technology, It makes a whole ton of sense to do.”
Walker brought up these comments during my visit, saying he disagreed. He said there’s a thirst for mobile manipulation and shared a story about one of the first companies interested in this new mobile manipulator.
“If you have a machine shop that has a bunch of CNC machines with a relatively high cycle time that don’t need to be tended very often, it’s really hard to justify putting an arm at every station,” he said. “If you’ve got one device that can serve multiple stations, then that makes sense.”
He added, “the things that have held customers back in the past has been how easy it is to set up, how much it costs and what it can do.”
“The reason we’re doing this product now is because we’ve helped solve the cost problem by having an arm and a robot that, over time, can get less and less expensive. We can integrate the systems more, we’ve massively lowered the entry price because we don’t have to have additional batteries.”
“This is an omnidirectional AMR with a 7-DoF arm, so you get something that’s actually better than the other solutions out there. Most of the other solutions are differential drive and a 6-DoF arm. So for me those are the key things – the energy and power, the degrees of freedom, the overall footprint capabilities, ROS integration, and ease of use. If this isn’t easy for everybody to use, then it’s a non-starter. Everything else falls apart.”
Editors Note: Portions of this story originally appeared on The Robot Report