Why not Industry 4.0?
“Industry 4.0,” caught on as a buzzword in 2011 when it was featured in the industrial strategy of the German government. The paradigm is best summed up by the ideal of a “lights off factory” where advances in cloud computing, robotics and artificial intelligence enable fully automated factories in which workers are optional.
Most industrial businesses operate according to Industry 3.0 standards today, with some larger enterprises on a conscious transformation from 3.0 to 4.0. Industry 3.0 refers to the revolution in manufacturing that arrived with computer numerical control (CNC) machines and other programmable hardware in the 1960s. Industry 3.0 rapidly advanced productivity on industrial teams, enabling the first ever industrial automation where human controllers could launch a process and a computer could see it through to completion.
Given the breakout returns of Industry 3.0, it was reasonable for evangelists of Industry 4.0 to foretell a similar opportunity. Yet despite all the publicity, most manufacturers in the US and abroad continue to follow the Industry 3.0 model.
Why is that? We on the Ascendant team have a hypothesis.
The short answer: bad data. To be precise, bad data & AI infrastructure. Some digital technologies like ERP systems can provide immediate lift to manufacturers. For the most advanced applications like robotics, the infrastructure to collect, parse and act on data repeatably at scale is only now emerging.
On the Ascendant team, we’ve spent the last several years building data-centric AI, engineering a blend of data infrastructure, data classification models, and machine learning operations (MLOps) technology to help clients like the US Space Force make effective use of automation. We’ve observed that the obstacles to implementing AI are usually grounded in underdeveloped data infrastructure.
Some of the limitations relate to a slow transition to the cloud-based data warehouses like Snowflake, but the problem extends beyond that. Most firms, including Fortune 500 companies and USG agencies, do not possess the mechanisms or the expertise to annotate the right data to build automated systems. One of the most important data inputs excluded from processing is input from subject matter experts. Right now, conversations around automation surface as an adversarial choice between people or technology. “Weak labels,” which is a machine learning term for input from experts are actually a fundamental requirement for building working automated systems. The implication? Working automation requires both people and technology.
At Ascendant, we’re sprinting past Industry 4.0 into Industry 5.0. For us, Industry 5.0 entails two things:
1. The promises of Industry 4.0, backed by mature data & AI infrastructure
2. Human-centric automation, where workers upskill to operate technologies that augment their productivity
We believe Industry 5.0’s emerging model of human-centric automation, perhaps better described as industrial worker augmentation, will win out over any efforts at building “lights out” processes. Industrial firms implementing Industry 5.0 technologies will be able to increase margins and increase revenue, paving the way for expanded valuation multiples. Adopting Industry 5.0 standards will also enable firms to ramp up contributions to causes that are critical to US competitiveness and the common good.
At Ascendant, we’re acquiring profitable industrial businesses and equipping them with Industry 5.0 tech. Invest with us or reach out for an always-free consultation on your succession plans. Together, we can share in the upside of building a resilient US industrial sector.
