The Future of Work: Smart Factory Evolution

Brian Byrne, Futurist and professor at York Schulich International, explains how humans + machines can work better, together. Based in Toronto, the Silicon Valley of Canada, he features three smart factory innovators that are humaneering the future, now.

Speculation about the future is always abstract until the arrival of a catalyst that propels speculative thinking forward into the present day. We are at that inflection point, as the long-lead global supply chains yield to localised systems within smart cities. The current trend of humaneering - combining AI + Human Intelligence, demonstrates how new jobs will replace automated work. Could machines eventually become so smart that humans are obsolete? Let’s look at three future factory innovators:

The rise of Kiva: Amazon’s race to arm robots for the future

In 2012, Amazon spent $775 million to purchase Kiva Systems to acquire “mo-bots” (mobile robots) that transport pods from worker to worker, using floor-mounted barcodes to navigate. Amazon now has over 200,000 of these machines working alongside hundreds of thousands of humans. This combined army has helped Amazon fulfill speedy deliveries to Amazon Prime customers. 

Kiva also provides foundation architecture to generate new versions of robots. Humans currently wrangle the robots but those jobs will shrink to fixers who will repair and maintain them. This has raised the bar on warehouse fulfillment, leading to the rise of other start-ups like Gravelle and Attabotics who build affordable robotic solutions to help smaller e-tailers keep up with Amazon. 

The USA warehouse industry currently employs more than 1.1 million Americans. Like transport drivers and autonomous trucking, this industry could be entirely automated. Mo-bots save human feet from the wear and tear of walking 12 miles a day on a hard concrete floor to pick and pack. Both stowers and pickers stand on cushion pads at their work-stations, holding a control device to guide machines. Amazon claims that they can store up to 40 percent more inventory in their warehouses, and that employees’ jobs are easier. 

Mo-bots enhance the average picker’s productivity targets from 100 items/hour to 400, as the New York Times reported in July. Yet, picking 3 times faster also implies fatigue due to repetitive, faster arm motions. The result is a shrinkage of rest time to mere seconds between each task, also complicated by 8-foot-tall shelving units that must be accessed by stepladders. This results in a workplace straight out of the famous episode of Lucy in the chocolate factory, where the worker is exhausted trying to keep up with the mo-bots increased delivery speed.

Here is where it gets interesting: robots are starting to achieve vision and motor control that are closer to human agility. Amazon has held an annual so-called picking challenge, to showcase agile robots. TU Delft Robotics Institute and the company Delft Robotics (both of the Netherlands) rose to the challenge: their robots selected items from a mock Amazon warehouse shelf at a speed of around 100 items per hour.

The UC Berkeley Robotics Lab made strides in this robotic arm race, a pair of humanoid arms controlled by a central system called Blue, performs agile work such as folding laundry using AI-powered eyes. Research lab OpenAI uses a technique called learning reinforcement to teach a robotic hand more precise and elegant movements. Kindred, a Bay Area startup, makes a robotic arm called Kindred Sort and is deployed in warehouses for the retailer Gap to blend humans and robotics to pick & pack apparel items.

Futurists foresee a time when NextGen robots will even repair and maintain their mechanical colleagues. All guided by a fulfilment AI brain which will be online and instantaneously transmit order flow, directed to the warehouse closest to the consumer’s address, somewhere in a drone-enabled smart city. From warehouse to front door, 100% machine delivered. All AI nodes would connect with each other via the cloud and coordinate resolution of individual warehouse out-of-stocks to ensure maximum fulfillment across the entire system. Humans will undoubtedly have new roles to play, but laborious picking and packing of items will likely not be amongst them.

Delegating assembly jobs to intelligent tech in the smart factory

Image credit: KUKA

KUKA AG is comprised of KUKA Robotics Corp. and KUKA Systems. Evolving from a 19th century acetylene gas company to Europe’s leading 20th century welding equipment firm was the first journey. The second was the firm’s shift into automating complex assembly and spot-welding work. Previously, spot-welding was a risky endeavour subject to defects and workplace injuries. In a KUKA Robotics factory, humans oversee robotic welders with a handheld control panel. Demo centres strategically located around the world enable customers to play with these control panels, while pre-testing robotic apps customised to their unique needs - all without disrupting their assembly operations back home. This decreases ramp-up time and preserves productivity.

For smart factory innovation, KUKA Systems offers bespoke solutions for holistic industrial production automation and component handling. Imagine filling a pit with automotive components and then watching robots sort, position and weld them together. BMW, Daimler, FCA and other automotive companies are loyal customers of KUKA, so bespoke solutions and training can be planned and optimised in co-operation with such clients.

Fast fashion using fast tech

AI is not yet adept at most creative endeavours with the exception of math-based elements such as chord progressions in music. The advent of fast fashion apparel over the last decades requires time compression of the cycle, starting with trend-spotting, designer prototyping, scale production and ultimate arrival in retail stores. Missing a trend by a few weeks can result in stacks of unsold merchandise to be marked down. Inditex, one of the world's largest fashion retailers and parent company of Zara, leverages assets that other competitors lack - physical Zara stores: 2,238 of them in 96 countries, each receiving shipments of prototypes and new items twice a week. 

Inditex has a 860,000-square-foot campus which houses 10 different factories; all of these manufacture Zara's most fashion-forward items of clothing - apparel that has special features such as pleating, ruching and insets. All factories sit next to the distribution centres and are connected by a network of underground tunnels that transport clothing on high-speed rails. Inditex has an innovation unit, led by former telecoms engineer Alejandro Ferrer and David Alayon. The unit tests ways of improving stock handling, partnering with California-based Fetch Robotics and Intel to test automated stock inventory and optimise matching volume of clothing to the right size boxes. 

Store to factory feedback loops

Inditex sends prototypes to select stores around the world which they know are harbinger locations - places where trendsetting urban super-consumers shop. These customers are invited in to try on hot fashion that is not yet in full scale production. Feedback is collected and transmitted via the web to the factory where refinements and scaling up production plans are made. Inditex knows that local fashionistas will give them the final opinions to produce apparel that clicks with other consumers.

Inditex focuses on location intelligence, ultrasound tech used to track footfall in stores. This intelligence allows apps to switch to instore mode when a person enters a store, so the customer can locate products and receive offers. It also provides data about clothing that is by-passed, where customers linger and which items they actually buy. The firm has also formed development partnerships with technology companies like Jetlore, which uses AI to predict consumer behaviour, and Spanish big data start-up El Arte de Medir. All data finds its way back to the factory floor in Spain. The result is a connected system that ensures retail front line data is transmitted in real time to the factory floor. Inditex combines the best human designers to create the fashion and integrates factory-meets-retail tech to deliver fast, at a good price, giving customers precisely what they want.

About the author

Brian Byrne is a trend forecaster, futurist and business war gamer who advises global corporations and speaks at major conferences worldwide. His team analyses mega-trends, emerging tech and consumer insights to design future scenarios so that clients can determine where to play and how to win. He teaches a “Future of Work” module at York Schulich international eMBA which hosts senior executives from European corporations and governments.