Autonomous Robots have existed in one form or another for more than 50 years. The first autonomous mobile robot – though hardly as functional as the ones available today – was developed at the Stanford Research Institute in 1969. “Shaky”, as it was affectionately known, was able to localize itself using machine vision, while its follow-on applications for automatic machine control have long been a benefit to manufacturing firms of all types.
In the last year, autonomous mobile robots (AMRs) have taken on added importance given the transmission risks related to human activity and COVID-19. Delivery robots, materials handling robots, even mobile disinfection robots have all been experimented with and more broadly deployed to rapidly fill the need for safe, sanitary, continuous services that previously could be provided by a human agent without a second thought.
AMRs are a great development, but only a first step in terms of the capabilities of autonomous robots. Their ability to engage in localization, planning and in some cases manipulation with less and less guidance is a fantastic development, but the complexity of skilled tasks – including the manipulation of industrial tools – cannot be ignored if robots are meant to fulifll their true value to both business and society as a whole. As such, value-added processes are an essential part of the field of autonomous robotics, and while AMR will grow vastly in the future, value-added autonomy will be more transformative to the world of manufacturing – and work – than even science fiction writers could have imagined.
How AMRs Have Progressed
Autonomous robots offer a few key benefits when it comes to automating mundane tasks like materials handling or mobile delivery. Their hardware costs generally provide a low barrier to entry, while the types of mechanisms used offer easy collaborative integration in a human-centric workplace (or world), and ultimately since they are required only to move between points that they identify on their own, they don’t require much of the intensive programming and scoping that comes with the standard robot integration.
As such, the ease of deployment and fundamental growth in the technology has led many analysts to expect the market to grow by 10 times between 2020 and 2030. A $220 billion market is nothing to sneeze ad, but in a situation with low barriers to entry, how soon will such an offering be commodified?
While broad growth in the sector is critical to freeing up human labor and amplifying productivity by other means, expectations can sometimes be overwrought. Consider, for instance, the promise of self-driving cars. 5 years ago, the promise was that full autonomy would be available now. Today, it’s rarely seen. To achieve road-safe, durable, easy to deploy autonomy, plenty of unexpected variables must be accounted for.
Ultimately, much of the expected growth in AMR is going to come in unstructured environments, but truly unstructured environments are no place for a robot – in part because they cannot adapt to absolutely anything, but also because they fundamentally lack human agency and ethical judgement, they are not suited to defensibly conduct themselves in a society where broad legal liability is a fact of life. In fact, environments where failure be reasonably accounted for by other safety measures are actually best suited to make the most of autonomous robots.
Understanding That Complexity Can Be Simplified
Ultimately, AMRs offer a straightforward technology for relatively simple jobs. The jobs are simple and valuable, but they are not necessarily value-added: that is, they are not transformative to whatever object they’re acting upon.
In that place, value-added processes can be taken up, but it needs to be understood what the difference is and why. A value-added process is any process or step in a manufacturing flow which transforms raw material into a more valuable good for a customer. Value is ultimately defined by a given processes value to the customer for a fixed good. In this case, it’s necessary to consider that different customers will have different needs, and ultimately this means each customer may have a different way of thinking – unless they’re willing to pay for the same thing.
In mass manufacturing, this is generally the case, whether it’s a car or a smartphone. That’s why mass manufacturing has seen robotics take hold so thoroughly – particularly in sectors like automotive or consumer electronics.
At the same time, high-mix manufacturers have served a greater variety of customers with big-ticket industrial orders. Sometimes these goods are destined for the consumer market, but most of the time they serve as the capital machinery that is essential to keep the economy going – transport goods, industrial machines, heavy equipment, large structural pieces or other uncommonly produced items.
A typical robot can’t adapt to these varied needs, and hence these sectors have relied primarily on skilled labor to get their jobs done. This is unfortunately coming to an end, but not by choice: there are more than 2 million unfilled manufacturing jobs on their way by 2030 in the US alone. China and other markets have also aggressively roboticized, and global population is expected to peak towards 2050, at which time a skills gap may still remain.
You see, in order to get to the point where you have enough robots to automate difficult jobs, you have to have workers produce a variety of essential goods which most of the robots found today cannot make themselves. Autonomous robots promise to change this: by giving robots greater adaptability for value-added processes, you can magnify the ability of skilled laborers that much more, and ultimately save on fixed costs in an advance almost as meaningful as the very first “Industrial Revolution”.
Value-Added Processes and Autonomous Manufacturing
Localization, planning and manipulation are the core priorities and tasks of autonomous robots. When it comes to value-added processes, process know-how is the fundamental missing link that can enable the core functions of autonomous robots to achieve new heights.
First, it’s important to understand how localization occurs with respect to a value-added process. Ultimately, the simples “tool” form of an autonomous robot can generate and execute its own process plan according to the location of a part while simply using a CAD file. At Omnirobotic, we provide a framework that allows for either a CAD file, 3D visual scanning technology or BOTH tools to identify the characteristic position and orientation of a part relative to a robot.
From there process know-how involves the ability to use an end-point tool – it could be anything from a paint gun to a welding torch or a buff finisher (at Omnirobotic, out technology currently supports spray processes) – and from there magnifies the capacity of this tool by taking in parameters whether it’s the intensity of the application or the width of the finish and then generates a motion program based on multiple goal sets that are needed to achieve a complete finished job.
In the past, it would simply be a human hand holding tool, with the arm and brain working to operate them. With an “Autonomous Manufacturing” approach, a tool can be considered not just the fixture itself, but the robot arm and its capacity to plan and execute a job based on the goal’s its assigned.
Technology and Social Progress
In recent years, the once unbridled optimism most of us have maintained in the benefits of technological progress have come under increased scrutiny. The pitfalls of digital technology, the app economy and its effects on social and mental well-being have become a hot topic of conversation with movies like “The Social Dilemma” and uncertainty about how powerful certain technology firms have become.
While information is a valuable commodity and our digital economy has given us access to it like never before, information alone can only take us so far. Pure software developments are attractive as both services and investments, but without impacting real-world hardware and processes, these kinds of advances will only have a limited ability to improve the actual material state of society.
We need systems that actually create for us, add value, function to fashion new goods, and not simply move them around or address secondary jobs aside from where the value itself is created. Autonomous Mobile Robots essentially have functioned as a utility up to this point – a service from point A to B, but not a tool. This is critical to understanding what value-added processes can gain by being roboticized: autonomous robots become a tool that does more, can be more, and are more useful for the human worker, augmenting our labor along with the actual goods we produce, instead of just moving them from point A to B.
With AutonomyOS™ and AutonomyStudio™, it’s never been easier to deploy an autonomous robotic system. Using 3D Perception with AI-based Task Planning and Motion Planning, manufacturing engineers and integrators can configure autonomous robotic systems for value-added processes that allow manufacturers to achieve more consistency and flexibility in production than ever before.