November 23, 2016      

Did you know that in the past six years, U.S. manufacturing has added 900,000 new jobs, even as adoption of automation increases around the world? Or that most industrial robots in operation have the intelligence of a 1-year-old?

These are just two of the many findings from the U.S. Robotics Roadmap, a 109-page document that was released on Oct. 31. With insights from more than 150 experts from around the world, the document is an attempt to develop a new robotics blueprint for the country.

The roadmap addresses many industry applications, including U.S. manufacturing and supply chain automation. But does it lay out a progressive vision for American robotics or simply restate what decision-makers already know?

Business Takeaways

  • Robotics in U.S. manufacturing and logistics is becoming increasingly “niche,” with a focus on how to manage different parts of automated solutions, like information exchange or interoperability.
  • Without a push to change the U.S. robotics strategy, businesses will likely see more fractured strategies at the local, state, and private level instead of a national vision (with the exception of military robotics).
  • The U.S. Robotics Roadmap, while a worthwhile read, needs to bring new ideas to the discussion about taking U.S. robotics forward.

Status quo for robots in factories, warehouses, classrooms

The U.S. is no longer the leader when it comes to designing and making robots, according to Henrik I. Christensen, a professor at the University of California, San Diego, and lead editor of the report. Instead, Japan, Germany, and Switzerland are leading.

The report stresses that, compared with the U.S., countries like China, India, South Korea, and Japan are taking robotics and artificial intelligence more seriously by investing in local industry and education.

For example, South Korea has invested $100 million in robotics every year between 2002 and 2012, while Japan plans to invest $350 million over the next 10 years in robotics.

At the same time, China and India are working to lure home students who are studying abroad — bringing back talent and making it work for them. The report mentions the U.S. need for more retraining, but it could be more specific.

According to the robotics roadmap, the U.S. lags in three areas: manufacturing, service, and medical robotics. They also represent some of the fastest-growing areas of the economy, growing at an annual rate of 9 percent and beating other sectors like media, toys, and home appliances.

Changes proposed for U.S. manufacturing and robotics

Using a timeline with milestones at five, 10 and 15 years, the report lays out a blueprint for advancing robotics in American industry.

Adaptable and reconfigurable assembly: The report sets a goal of being able to produce goods such as vehicles in the shortest span of time possible.

Fifteen years from now, an entire assembly line should be able to be customized within one hour to meet changing requirements. Currently, it can take up to two years to alter an assembly line for a new car model.

Dexterous manipulation: The roadmap states that the fingers on a robotic arm should work as well if not better than human hands.

It projects that within 10 years, certain robots will possess “whole-hand grasp acquisition” and a minimum level of dexterous manipulation. In other words, robotic arms must overtake the dexterity of human hands.

Model-based integration and design of supply chains: To fuel a new kind of supply chain, the report calls for more investments into interoperability between different systems. This extension should improve the exchange of data among components, dovetailing with the concept of the industrial Internet of Things.

The blueprint predicts that within five years, new manufacturing locations will be created without flaws, which it calls “bugs.”

The roadmap also mentions nano-scale manufacturing (a step beyond 3D printing), the importance of robot-to-human safety, and developments in autonomous navigation as enablers of next-generation U.S. manufacturing and logistics.

Leading or following?

Most of the time lines and accompanying objectives do not state anything groundbreaking but instead reflect what many observers already see as the natural progression of robotics.

Some of the roadmap’s goals are underwhelming. For example, it expects that within 15 years, autonomous vehicles will be able to operate in any environment where humans can. Isn’t that an obvious and implied objective of the self-driving car industry?

In addition, the document predicts that, within 15 years, robot arms will be “increasingly common partners with humans in manufacturing” (p. 80). Aren’t collaborative robots already doing this in factories and warehouses worldwide?

Multi-agent robotics

The 109-page report also touches on multi-agent robotics, which is the ability to deploy multiple robots to work together to meet objectives, instead of relying on a single robot. This has major applications for both U.S. manufacturing and logistics, and the report lays out some fundamental challenges that need to be overcome.

For example, when it comes to human-robot interactions (HRI), there is still a gray area about the best way to supervise numerous autonomous machines in an operation. Most providers of robotic fleets today provide management consoles or dashboards along with proprietary software.

Should there be a central platform that allows the human controller to make changes, an AI managing the robots, or something else entirely?

Alongside HRI concerns, the roadmap raises the question of how robots should share information. Does there need to be a central platform that acts as an “information hub” that all the robots access, or should data be localized (stored on each robot) to ensure that the robots can continue operating if the hub or other robots stop working?

In the case of mobile robots in hospitals, delivering packages to homes, or processing payments in retail or custom manufacturing, privacy and security concerns are also relevant.

As with industrial automation and U.S. manufacturing, the report sets objectives that are predictable extrapolations.

For instance, the document predicts that multi-agent data sharing in 15 years will simply rely on the “Internet of Robotics Things” (IoRT). Beyond the obvious definition, what does this actually mean?

More on U.S. Manufacturing and International Robotics Policy:

A need for disruption

Although the U.S. Robotics Roadmap came out a week before the presidential election, it failed to lay out a truly disruptive vision for U.S. robotics leadership that the next administration could develop or follow.

With no “moon-shot” objectives, the report appears to be more of summary of industry progress than a new strategy for U.S. manufacturing and automation.

Another problem is that if the roadmap received input from 150 experts, whom else should businesses turn to for government leadership and out-of-the-box thinking? Are traditional avenues such as academic and engineering laboratories enough, and what should we look for from innovation labs and startups?

The U.S. Robotics Roadmap sets many objectives for U.S. manufacturing and supply chain automation in the coming years, but it offers few ways to bring new ideas to the discussion. It doesn’t go into great detail on economic policy, fostering stronger ties within the robotics ecosystem, or international cooperation.

If you’re part of a business or government looking for general guidance around industrial automation, the U.S. Robotics Roadmap is a worthwhile read. But if you’re looking for new opportunities for robotics and AI, it may not be the blueprint you were hoping for.