Latest news with #Unimate
Daily Maverick
27-07-2025
- Science
- Daily Maverick
Crossed Wires: The robots are coming, the robots are here
Robots are quickly colonising every arena of human endeavour that requires physical labour or dexterity, and the convergence of new generative AI models and robotics is going to supercharge the industry. Science fiction fans will remember Isaac Asimov's prescient 1950 short story collection, 'I, Robot'. The stories in this iconic collection revolve around the memories of robopsychologist Dr Susan Calvin, 75 years old in 2057. There are many ethical themes in his stories that mirror the hot-button technology debates raging today, but Asimov's futuristic vision appears to have been a couple of decades off. The robots aren't coming some time in the future. They're already here. It is sometimes difficult to pin down why a particular subject bubbles to the top of the news cycle. AI, of course, has taken hold of the headlines and won't let go; its grip is fierce. But robotics has been around for a very long time, since the first commercial industrial robot was developed in 1961 (the 'Unimate' at General Motors), and it's been around even longer in the human imagination. The first point of interest is that there is a profound change happening in some sectors of robot production. The original robots, most of them built for industrial manufacturing, were essentially a collection of servo motors, spherical joints, pincers, cutters, drillers, welders and the like, all operating under very precise instructions, repeating the same physical actions ad infinitum — or until their instructions were modified in line with changing production requirements. This description somewhat simplifies the intelligence built into these robots, but the key point is that the instructions for physical actions over time were predetermined and cast in silicon, driven by hardened computer code. As the technologies of vision, touch, movement, location awareness and proprioception have advanced, so have the robots undertaken more ambitious (and sometimes audacious) jobs, such as critical surgery in an operating theatre. All with increasingly exquisite sensitivity of both fine and gross motor control. This brings us to the question of which countries are currently on the robotics playing field. The US, having outsourced nearly all its manufacturing requirements during the heady days of globalisation, didn't even try to take a leadership role in robotics. Surprisingly, China is not leading either — it became serious about robotics rather too late (around 2015). Until recently, the top 10 robot manufacturers have been Japanese (8) and German (2). In a remarkably unChinese move, a company called Midea Group, based in Guangdong province, acquired the German company Kuka in 2017, clearly taking note of the capitalist tactic of buying one's way into technology leadership rather than doing the hard work of building and competing. They're doing everything What are the latest machines doing? Increasingly, everything. Manufacturing, obviously, both heavy and light. Add medical robots doing everything from brain to spinal surgery; nano-robots of less than 100nm in size for targeted drug delivery (in pre-clinical development); agricultural robots for planting, harvesting, plant healthcare and packing; and military robots with a bewildering and sometimes scary array of offensive and defensive capabilities (see below — a rather alarming picture of Ghost Robotics' robodog Spot mounted with an automatic weapon). In short, robots are quickly colonising every arena of human endeavour that requires physical labour or dexterity. This leads to the question of how the robots are performing. Are they more productive? Are they taking jobs? The data are startling. Dispiriting for some and exciting for others, depending on what you do to earn your living. As with all transformative technologies, it's a mixed bag of pain and pleasure for those caught in its net. Here are some statistics: A 40× increase in global robot stock since 2000; A 15× increase in robots per worker since 2000; A 30% increase in productivity compared to human labour by 2030 (McKinsey); A 90% reduction in manufacturing defects (Foxconn iPhone production); A 3× decrease in the return on investment period since 2000 (down from 10 years to about three years); A 35% increase in crop yields (forecast); and A total of 85 million jobs lost by 2035, and only 20 million created (World Economic Forum). There is more, but the picture is clear. There is no stopping this train. To return to the profound change mentioned earlier in this column, the convergence of new generative AI models and robotics is going to supercharge the industry. The core foundation of robots following a precise set of instructions (albeit often complex) is being reshaped. Robots are now being built that can learn autonomously and continuously from their environments (sight, audio, touch, 'smell', heat, humidity). They can be addressed via vernacular human speech, learn from their mistakes, communicate to solve problems with other robots and access vast stores of knowledge now available from companies like OpenAI. (For anyone looking for a glimpse of the future, watch this 2.5-minute video; take note of how the robots communicate with each other.) At this point, the ghost of Isaac Asimov might raise its head. We are already on the edge of AI systems that can set their own goals. We have already seen indications of deceptive behaviour by these systems, in both controlled and uncontrolled experiments. Bad behaviour and misinformation (such as Grok's racist outbursts) are now part of the AI landscape. Asimov's famous three laws of robotics come to mind: protect humans, obey humans, protect yourselves. They were followed by his 'Zeroth Law', which updated and replaced the others: A robot may not harm humanity, or, by inaction, allow humanity to come to harm. It's a nice thought, but I am not sure we know how to build that into the great robot revolution. And, even if we did, it is probably too late. DM (For an in-depth but concise look at the robotics industry, check out this article.) Steven Boykey Sidley is a professor of practice at JBS, University of Johannesburg, a partner at Bridge Capital and a columnist-at-large at Daily Maverick. His new book, 'It's Mine: How the Crypto Industry is Redefining Ownership', is published by Maverick451 in SA and Legend Times Group in the UK/EU, available now.
Trade Arabia
13-03-2025
- Automotive
- Trade Arabia
The factory reset that global manufacturing needs
As manufacturers contend with onshoring policies and the impact of extreme weather events, industrial AI applications will be essential to maintaining costs, unlocking production efficiencies and staying competitive, says Jim Chappell, Global Head of AI and Advanced Analytics, AVEVA Visiting a car manufacturer in Dearborn, Michigan, Japanese executive Eiji Toyoda came away with a startling insight to stockpiled inventory: What if automobile parts were produced just in time for use? That brainwave was first adopted at the Toyota Motor Corporation in the fifties – a company Toyoda went on to lead. In the wake of the 1973 oil crisis, the concept became a defining feature of the Japanese car industry. By minimising waste while improving quality, the production system garnered Japanese cars a reputation for reliability and affordability. Just-in-time assembly, often called lean manufacturing, has since become a bedrock of global management canon . Its adoption in the seventies and eighties occurred amid periods of economic stagnation and high inflation, as companies battled to improve business margins and slash production costs while doing more with less. From 1973 to 1984, Toyota alone grew production from 170,046 units to 3.4 million units, the company's records show. The rest of the sector was quick to follow suit. Risks cloud forecast for manufacturing Manufacturers around the world face a similarly transformative moment today — although the world is in a very different place. Covid-linked supply chain issues may have eased over the past few months. However, new risks cloud forecasts not just for the immediate term, but well into the next decade, as the World Economic Form noted in its rather bleak Global Risks Report 2025. Geopolitical turbulence and the emergence of a multipolar world are increasing demand for onshoring production, while new trade tariffs will have their own impact. Extreme weather events, meanwhile, will continue to weigh on the supply of raw materials, including both ingredients and mineral resources. But when the going gets tough, to reframe the old adage, the tough turn to new tools. In every economic crisis, manufacturing has leveraged cutting-edge technologies and processes to buck economic headwinds and transform business outcomes. The technology for our times is industrial AI, and the true gamechanger will be its ability to generate real-time insights. Or just-in-time insights, to borrow from Eiji Toyoda. Leading from the front with business transformation From Henry Ford's embrace of the assembly line in the early 1900s to General Motors' Unimate and Toyota's just-in-time and Kanban systems, the manufacturing sector has led the pack in terms of innovation and business transformation. The advent of Industry 4.0 has speeded up digitalization and helped launch autonomous plants. But as manufacturers face unprecedented disruptions, first-mover advantage in the current climate appears to rest with those embracing the role of industrial data and related technologies in the artificial intelligence (AI) family, such as machine learning, predictive analytics and generative AI. Together, they are critical to providing the insights needed to keep manufacturing operations lean and efficient. It's fair to say the real battle won't be offshore vs onshore—it'll be between smart factories and outdated operating models. That's something production teams at Barry Callebaut have experienced firsthand. The Swiss-Belgian manufacturer began taking a digital approach to making chocolate seven years ago, using advanced industrial software products to create a smart factory that brings together people, processes and technology. By integrating enterprise visualization with a next-generation manufacturing execution system (MES), the chocolatier improved traceability and boosted productivity. With this connected digital backbone, it has eliminated data silos and empowered staff with real-time insights from across the value chain. Alongside, predictive models have revealed optimization opportunities for instantaneous adjustments, increasing production capacity by 10%, while system-wide efficiencies have cut energy use and put the consumer products leader on track to achieve its net-zero goals without sacrificing output. To stay with the chocolate industry, Nestlé used advanced AI-infused analytics and real-time data in the cloud to ensure consistent flavours across each jar of its Nesquik family drink, while cutting wasted powder by 10%. At 101g saved per 1kg jar, that's 10 extra cups of Nesquik. The company now wants to expand that success to more plants and to other products such as Ovaltine. In the US, meanwhile, New Belgium Brewing Co. has gone from craft brewery to national leader with a digital-centric approach, using a MES platform together with advanced operations control software. Thanks to improved operational scheduling, process visualization, advanced AI analysis, and better collaboration and digital knowledge sharing, the Colorado-headquartered company has streamlined production while boosting efficiency and quality. Within just two years, overall equipment effectiveness rose from 45% to 65%, while downtime dropped 50%. Further, despite local variables, beverages produced at different facilities across the USA now taste identical. Majority of manufacturers demand new technologies As more companies look to unlock such transformative gains, 58% of manufacturers say the need for new technology to empower their workforce is a top business challenge, according to the AVEVA Industrial Intelligence Report 2024. Indeed, the overwhelming majority (97%) believe industrial AI solutions and other digital technologies are required more than ever to remain competitive in today's challenging landscape. Yet, the majority of digital transformation projects – 78% according to Capgemini research – fail to deliver their promised benefits because of poor alignment with business outcomes, limited visibility across end-to-end operations and sub-optimal insights. The answer lies in implementing flexible and open systems that bring together distributed enterprise and operations teams around a comprehensive digital data thread, where they can view just-in-time business insights at the scale they need.
