This year’s EmTech Next, a future-focused conference of innovators from industry, government, and academia, considered the future of work. The meeting posed a simple question: are we ready for a workplace shot through with AI and Robotics? If you were to judge from the program, the real challenge ahead will be developing policies for a world characterized by autonomous factories and intelligent machines. We have to be ready to define the place of humans against the inexorable creep of automation. The message appeared to be one of preparation.
The picture the speakers painted was a little different. Intelligent automation is here, but it has its limits–notably cost, scalability, maintenance, and flexibility. AI gets smarter every day, but algorithms aren’t ready to replace human minds.
The world the speakers envisioned was one of human-machine collaboration, not conflict. Of collective intelligence, not human complacency in the shadow AI superminds. Of machine augmentation of humans, not replacement.
As co-host Erin Winick of the MIT Tech Review noted, for a conference devoted to AI and robotics, humans were conspicuously front and center.
The future of work is a topic particularly relevant to manufacturing. Tulip CEO Natan Linder’s talk addressed the place of humans in the industry. His argument is relevant to all manufacturers.
Bridging the Skills Gap
Manufacturing is facing a massive labor shortage.
Some of the reasons sound familiar: Industrial automation now performs most repetitive tasks. Shop floor jobs are hard and boring, so turnover in manufacturing is high. And the half-life of skills in manufacturing is short.
But the truth is a little different.
In reality, manufacturing faces a labor shortage because there aren’t enough workers who can do the sophisticated, creative work the modern factory work requires. Those who are capable often lack interest in the industry.
As Linder noted, “If you go in front of a classroom of university students and you say, ‘who wants to work in manufacturing?’ No one raises their hand. If you say, ‘who wants to work in robotics, data science, and analytics?’ Everybody raises their hands.”
And yet these are all skills required by the modern manufacturing engineer.
This skills gap, this lack of alignment between the abilities required to succeed in modern manufacturing and the abilities and aspirations of the workforce, is estimated to cost $2.5 trillion over the next decade.
As experiments in wholesale automation fail to deliver results, the question has become, “How do we attract qualified workers into manufacturing?”
The State of Manufacturing
The answer to this workforce crisis isn’t a PR campaign. The problem runs deep, and it’s going to take a concerted program of education and collaboration to resolve. Most importantly, it’s going to take a redefinition of work in manufacturing.
As Linder argued,
“It’s not about making manufacturing sexy. It’s about making it fit the way we work.”
This last point–“making it fit the way we work”–is essential. So what does that mean?
Over the last twenty years, software has come to dominate our personal and professional lives. Entire professions have been redefined by the software tools available to them. There are often entire tech stacks designed to empower workers in their roles.
These software tools are user friendly. Crucially, they’re generative. That is, software that lets workers create more software, either through application development, integration, or some other means. To a certain extent, we’re all developers now, and job functions have changed to reflect that fact. For example, marketers are now responsible for brand management and analytics, previously separate domains. Account executives now prospect and sell as well as administer their database of contacts.
So it’s significant that this is not the case in manufacturing. If you walk out on a modern shop floor, you’ll still see clipboards. You’ll still see stopwatches and paper forms. Where there is manufacturing software, it’s often rigid, expensive, hard to update, and difficult to maintain. Traditional manufacturing software operates on a timescale too slow for an industry facing rapid NPI cycles, short product life-cycles, and demand for customization.
Further, the lack of flexible manufacturing software hasn’t prevented job descriptions in manufacturing from changing over time. Now, manufacturing engineers are expected to do work previously performed by software developers and data scientists. Simply put, the tools in manufacturing haven’t kept up with the challenges of the job.
The Way Forward
This fundamental insight–that the manufacturing engineer’s tools have not evolved with the job–offers a way forward.
If it were as simple as writing software, though, there wouldn’t be a problem.
Part of the reason manufacturing has lagged behind other industries boils down to a single factor: manufacturing is physical. There is no manufacturing with about machines, materials, and labor. Engineers need to be out there on the shop floor working with physical processes, not cloistered behind a monitor.
Any software tool thus needs to account for the physical reality of the factory by relaying between back-end system and front-end processes. It needs to choreograph machine and human activity in a truly sophisticated way.
What Linder proposes, and what we’ve built at Tulip, is a Manufacturing Application Platform that lets engineers design solutions to the challenges they face every day.
As he described it, “We complete the last mile of shop floor automation by giving manufacturing engineers a self-service tool that turns every station into an instrumented, data collecting, guided environment enabling continuous improvement.”
This platform reflects the fact that engineers are expected to write software by allowing them to design apps without writing a single line of code. It reflects the fact that the manufacturing engineer is responsible for handling large data sets by streamlining data collection and analysis, and organizing data in custom tables accessible without writing SQL.
Most importantly, Tulip makes a new kind of engineering work possible, one suited to the reality of the modern shop floor.
All of the exciting, futuristic developments noted at EmTech Next are integrated into the platform. Tulip has AI analytical and computer vision systems; it connects devices and machines through IoT; it stores data and computes through the cloud; it automates several previously manual process.
But it does so by recognizing that human work is central to the success of every one of these features. When engineers are empowered with the right tools, the possibilities for improvement are endless.
If manufacturing is going to close the skills gap, then it needs to focus on augmenting workers with the right technologies, not replacing them. The industry needs to find a way to empower engineers as knowledge workers, rather than impeding them.