MOTUS LABS M-DRIVE BRINGS VALUE TO THE ROBOTICS INDUSTRY
| By: Team Motus
TODAY'S ROBOTS ARE SLOW, AND THEIR WORKSPACES ARE LIMITED
This statement may seem somewhat shocking. But while there have been tremendous advances lately in fields such as machine learning and vision systems, most articulated robots cannot move much faster than the earliest robots introduced in the 1950s. Why is this?
The simple answer is that components are not available to make truly lightweight robot arms. Some progress has been made in lightening the links between joints through the use of composite materials, but the active components in robot joints - actuators - have not fundamentally changed for several decades. There is so little innovation in this area, that robot designers do not even acknowledge that there is a problem. A problem without a solution is no longer seen as a problem, but rather a constraint. Designers simply see no way around it.
The hidden costs of these shortcomings can amount to thousands of dollars. Based on the ROI model used by the Robotic Industries Association, even a moderate increase in speed can result in an added lifetime value equal to or exceeding the cost of the robot.
ROBOTIC ACTUATORS - THE MUSCLE BEHIND ROBOTS
The actuators used today in articulate robot arms, that is arms with six or seven joints, usually are a combination of a motor and a drive/transmission. Actuators are generally described in terms of the amount of torque they are able to produce. Torque is measured in Newton-meters (Nm) and can be understood as a “turning effect” that produces a rotational force. Each actuator is responsible for moving the portion of the robot arm ahead of it. It is easy to see that the further away an actuator is from the end of the robot arm, the more weight it must move. The conundrum in robot arm design is that the more weight an actuator must move, the heavier it becomes itself. Since each actuator down the arm is moving more weight, each successive actuator becomes heavier and heavier, imposing its own weight burden on the next actuator above it. A small to mid-size robot might require actuators able to supply between 20 and about 400 Nm of torque. Larger robots might require actuators able to provide thousands of Newton meters.
TORQUE DENSITY IS THE KEY TO ROBOT PERFORMANCE
An important measure of actuator performance is something called torque density. The torque density of an actuator is equal to the amount of torque it is able to deliver divided by its mass. If, for example, an actuator is able to produce 50 Nm of torque and weighs 5 kg we would say that it has a torque density of 10 Nm per kg.
An actuator usually includes a motor along with a transmission. The motors used in robot arms are able to produce small torques at very high speeds. The purpose of the transmission is to convert the low torque, high-speed energy of the motor to a much higher torque at a lower speed. The gear ratio of the transmission can be thought of as a torque multiplier. For example, a transmission with a 100:1 gear ratio will step up the torque of a 1 Nm motor to 100 Nm. At the same time, the transmission output would spin at 1:100th of the motor's input speed. For this reason, robot transmissions are sometimes also referred to as speed reducers.
"My intent is to revolutionize the robotic actuator industry to enable
new applications and breakthrough old technology limitations."
Carlos Hoefken - Inventor & Co-Founder, Motus Labs
STAGNANT DEVELOPMENT & CURRENT LIMITATIONS OF TODAY'S DRIVES
Contemporary robot transmissions operate on basic principles of gearing that date back to the 5th century BC. In 1957, The American inventor C. W. Musser patented a novel new type of gearing that featured a flexible steel element. Dr. Musser's invention, alternately called strain wave gearing, increased the area over which the drive input and output elements meshed. At the time, this was a significant innovation, enabling the output load to be distributed over a larger area than before. This provided for higher torque
densities than previous gear drives were able to attain.
Harmonic and Nabtesco drives allow for around 15 to 20% engagement between input and output gear surfaces. With the stain wave drive, however, spur type gear drives seem to have reached a sort of plateau in torque density. Any incremental gains made in torque density since the strain wave drive's introduction have usually been at the expense of compromising other parameters, such as stiffness or efficiency.
"Torque to weight ratio is the key. "
Esben Ostergaard - Co-Founder, Universal Robots
THE MOTUS LABS BREAKTHROUGH
The Motus M-DRIVE technology provides a radical new means of reducing actuator weight through a new type of drive/ transmission. In order to achieve greater improvements in torque density, a revolutionary new design was invented by Carlos Hoefken. The Motus M-DRIVE enables up to 80% engagement of input and output. Strictly speaking, M-DRIVE is not a gear drive in the conventional sense. Unlike a strain wave drive, the input surfaces of the M-DRIVE are distributed outside rather than inside the output element and they are set in motion by means of a cam follower mechanism. This arrangement allows the input load to be distributed over 4-5 times the area that is achievable in a strain wave drive. This in turn enables a correspondingly higher torque density: up to four times the torque in the same weight. To date, Motus Labs has been awarded seven U.S. patents for this novel technology. The mechanical advantages of the Motus design also offer potential improvements in other performance areas – most notably efficiency and torsional stiffness.
Strain Wave drives produce a lot of heat due to their use of flexible metals – with internal temperatures approaching the limits of what the unit's lubricant is able to sustain. As a result, strain wave drive manufacturers specify hard limits of 90 minutes or less to the time during which the unit can operate continuously. The excess heat generated by strain wave drives also poses challenges for motor suppliers: in order to sustain reasonable motor performance, a safety margin of 20% or more must be assigned to the motor's rated torque. As a result, robot joint designers are forced to purchase more motor torque than they really need and suffer the resulting weight penalty. In contrast, M-DRIVE architecture is designed to run cooler. By using the Motus Labs solution, a robot designer can avoid this situation.
Stiffness, or torsional stiffness, describes how much a gear drive deforms when bearing a load. A well-known weakness of strain wave drives is their elasticity, which can introduce mechanical resonances and instabilities at low speeds. Since torsional stiffness generally improves with drive size, a common workaround today for robot designers is to use a drive with more torque than they really need – again introducing unwanted weight and cost. The Motus M-DRIVE design provides higher torsional stiffness because of increased mating surface contact area eliminating the elasticity.
FLEXIBILITY IN MATERIALS
Due to the nature of the design, Motus drives are built with standard alloys and require no precision machining. Motus drives have been built using steel, aluminum, and have been 3D-printed in plastic. The flexibility in materials and simplified manufacturing requirements translate directly into cost savings for Motus' customers, as well as the ability to address unique applications. Motus has been called on in the past to deliver aluminum versions of its drives for drone applications, as well as for applications where non-magnetic materials are needed – such as semiconductor handling and MRI machines.
WHITEPAPERDIRECT ECONOMIC BENEFITS FOR ROBOT MAKERS
When a robot is designed using Motus M-DRIVE, a number of benefits accrue:
- Motors can be “right-sized” without the need for excessive safety margins to accommodate gear drive heat.
- Beginning at the wrist, each actuator needs to provide less output torque – resulting in progressively lower weight gear drives at each joint.
- The need for each actuator to provide progressively less torque lowers the torque requirement for each motor – further driving down cost and weight.
As an example, a typical low payload (0-20 kg) robot may see the following benefits when implemented with a Motus solution:
- Up to 50% less torque required per actuator
- Up to 65% less weight per actuator
- 10% to 20% lower cost per actuator
This does not include any additional savings due to “right-sizing” the drives for torsional stiffness – perhaps leading to an additional 10% to 20% cost savings per drive or actuator.
ECONOMIC BENEFITS FOR ROBOT USERS
Motus' advantages allow robot makers to better serve their end customers through:
• Larger robot workspaces through increased reach
• Greater productivity through increased speed
• Longer robot life through decreased wear
Depending on the robot type and installation, the improvements afforded by MOTUS M-DRIVEs can offer savings to the end-user equal to or exceeding the cost of the robot. Where do these savings come from?
The MOTUS M-DRIVE can increase the torque density of the wrist actuators by a factor of two – meaning that their masses decrease by half. This implies that the arm length can increase up to 68 cm – nearly 40% longer. Longer reach means that the robot work cell can be larger, potentially reducing the number of robots required to service a given user facility.
Robot arm speed and acceleration depend directly on the amount of torque each actuator is able to supply to move the weight of each link and joint. Here the conundrum is that the more torque an actuator must supply, the heavier it becomes. As a result, the weight of each successive actuator becomes a drag on the ones that follow it. A reasonable analogy might be trying to run with weight belts on one's knees and ankles.
The MOTUS M-DRIVE torque density advantage gives the designer a large safety margin for reliability. Whereas, for example, a conventional Size 17 gear drive might provide around 30 Nm of torque, the Motus drive could provide up to twice the amount of torque in the same size. This additional torque “headroom” translates directly into less drive wear and lower maintenance costs. These results may vary depending on the robot geometry but in general, the robust torque ratings of Motus drives will result in a 20-50% increase in the useable life of the actuator.