$150 Billion! That is how large the ‘Industrial Automation” market is today. In fact, in a span of 150 years, we have invested several trillion dollars as capital towards industrial automation. What returns did society reap?
Between 1AD and 1800AD, world GDP grew only as much as the population grew. GDP relied solely on human hands. This implies that the per capita GDP, and thereby individual wealth and standard of living, remained poor.
The 2nd Industrial Revolution
Intriguingly, the world GDP took a dramatic turn, coinciding exactly with the advent of the 2nd Industrial Revolution circa 1870. By 1973, through a century of mechanized automation, the world population had grown by 300%, but the GDP had grown by a whopping 1400%. Even this massive feat pales in companion to what has been achieved since the 1950s, when large scale automation began, and the silicon revolution unfolded. By 2000AD, while population had grown 4.6 times since 1870, GDP had grown over 30 times in the same time frame. Consumerism soared and the middle class boomed. What industrial automation did to alter the course of our evolution will remain a crucial chapter in the human story for ages to come.
That automation, even as small as a tool upgrade, propels productivity is intuitive to discern and extrapolate through simple observation. A worker enabled with a power screwdriver can drive 6 times as many screws per hour than one without. We can only imagine the productivity boost through inclusion of a robotic arm – the pinnacle of such power tools.
It is no wonder that the per-worker manufacturing productivity is disproportionately higher in advanced economies that have invested massively in industrial infrastructure comprising of such tools. In comparison to India, although it costs 2x as much to hire a manufacturing worker in China, 5x in Japan, 7x in Germany and 8x in the US, manufacturing GDP per-worker in China is 4x, in Germany and Japan 16x. and in the US a whopping 32x that of India (Figure 2).
Despite these benefits, industrial automation adoption is still in its infancy. In contrast to the $150 billion spent on industrial automation, the US alone spends around $1.3 trillion on manual labor. The market for industrial robots and automation market is projected to grow at 8-10% CAGR. At this rate, it will take over 50 years for global annual automation investments to just match what the world spends today on manual labor. The internet economy in contrast has been projected to grow at 37% CAGR. If only we could deploy robots as fast as we adopted internet apps like Amazon or Uber, the scale of impact on our economy’s growth would be unprecedented.
Automation Post Covid-19
As we are seeking to rebuild our economy post COVID-19, automation has now emerged as a pressing societal necessity, rather than a business advantage. However, the recent disruptions that brought manufacturing to a grinding halt, even in heavily automated industries, raise alarm. Trillions of dollars of industrial equipment were lying idle for months even as the world struggled to produce and supply essentials goods in high demand. Elon Musk’s inability to produce and supply ventilators at Tesla stands testimony.
In our quest for efficiency and productivity, we ignored versatility and adaptability, and thereby resilience.
A physical task is all about manipulating objects – picking, orienting, and placing objects in an expected orientation and location. Yet, a $18,000 – $30,000 robotic arm which can be pre-programmed to move thousands of parts a day with as high a placement accuracy as 20 micrometers, is easily incapacitated from functioning if the object is misplaced by even as low as 0.2mm.
Ensuring such a rigid Object Structuring demands a sophisticated network of sensors, mechanical contraptions and custom conveying systems which incurs months of design, development, integration and testing. This drives the total robot deployment costs as high as $100,000/- to $150,000/. It takes 2x-6x the cost of a robot to just arrange an object in a way, so that the robot can pick it blindly.
All this effort and expense is rendered useless if the object’s dimension changes even by a millimeter, let alone replacing the entire object. On the contrary, humans can effortlessly adapt to such changes and beyond, in spite of the human arm’s inferiority in its precision, speed and power in comparison to a robotic arm. The robotic arm might be the pinnacle of re-configurable power tools, yet it is just a tool. Whereas humans are armed with a sophisticated faculty of Vision, Manipulation, Proprioception, and an unparalleled Visual Intelligence purpose built for Object Manipulation – an easily trivialized & barely explored subject.
Product Specific Factories
This insistence for rigid and costly object structuring infrastructure by robots has limited their adoption only to manufacturing sectors such as automotive, who have a long and predictable production lifetime and large volumes that justify these upfront costs. Manufacturers of rapidly evolving products like smartphones, high variability goods like apparel and low-volume, high-complexity products like aircraft, satellites, and medical devices, have barely adopted robots especially for assembly tasks. The factors described above have resulted in Product-specific Factories, as we know them today .
Post COVID-19, if we are to rebuild our industries with focus on resilience, while continuing to harness the scale, speed and GDP growth associated with automation, we need to re-imagine our robots. Robots must cease to be task specific and acquire the cognitive ability to see, understand and learn to manipulate all kinds of objects, even if they are cluttered or obstructed. This lack of visual ability for robots to perceive and manipulate a variety of objects must be addressed through significant technological breakthroughs blending machine vision and artificial intelligence.
I believe the future will witness the rise of product agnostic Universal Factories armed with assembly lines of visually intelligent robots that can rapidly morph to produce various types of components or products on demand. A smartphone assembly line today, a healthcare PPE line tomorrow, a rocket engine nozzle line day-after. Anything less, limits the true potential of robotic automation.
About the Author
Nikhil Ramaswamy is the CEO of CynLr, a VC backed deep-tech robotics startup enabling robots with visual intelligence to manipulate objects. Prior to co-founding CynLr, Nikhil worked with co-founder and CTO, Gokul NA at National Instruments, a billion-dollar MNC in the test, measurement and control automation space, where Nikhil served as a Key Accounts Manager and Gokul served as a Specialist in Machine Vision, Embedded and RF Systems. He can be reached at nikhil[AT]cynlr.com.
- Automation Imperative Accelerates
- 10 Lessons Learned About Robotics and Automation in the 2010s
- Case Study: Why Ford Deployed AMRs to Automate Spanish Factory
- Nord Modules Enters U.S. Market to Support AMR Adoption
- Vector AMRs From Waypoint Robotics Bring Accuracy, Ease of Use to Factories
- Warehouse Automation Companies – The Top 50 for 2020 According to LogisticsIQ
- Seegrid Surpasses 2M Miles Driven Safely for its Autonomous Vehicles