As travel begins its recovery, the industry continues pursuit of a world where in this decade, thousands of air taxis may fly over our cities. But there are still large technical, regulatory, and societal challenges that must be solved to make this vision a reality.

Air taxi programs are being established by aerospace titans and startups. Both are using hardened aerospace engineering practices, adopting existing regulatory frameworks and building in the same safety concerns as other aircraft.

How does an industry accustomed to complex manufacturing in low volumes transform to high-scale production? Fortunately, through huge advancements in cobot technology and artificial intelligence (AI), we can address this concern and begin making air taxis a reality.

Manufacturing an air taxi

From a manufacturing perspective, most air taxis will be built leveraging existing processes and through the existing supply chain. But how do we scale a historically low-scale industry that has — in large part — not made the investments in digitization and automation that have transformed adjacent industries, such as automotive?

Some aerospace manufacturers have considered lifting-and-shifting the automation used in the automotive industry, but this approach is problematic for several reasons.

  1. Most aerospace concerns don’t have the time or capital to invest in the large-scale automation that powers today’s most advanced automotive factories. To do so would require re-architecting countless production lines.
  2. Even if they did automate a new production line based on an automotive-industry template, typically volumes are not high enough to recoup the up-front expenditure.
  3. Most automation to drive scale in the automotive space is physically built to handle a single or — at most — a very small set of functions in an assembly line, with support for a limited number of configurations. Given the vastly smaller volumes in aerospace manufacturing coupled with more complex configuration, such single-function automation simply can’t scale cost-effectively.

Cobots and AI are the way forward

The answer is to introduce high-scale automation at low cost and — critically — at low disruption. Ultimately, the automation we devise for aerospace manufacturing needs to work and drive value at high, low and even no scale. The combination of cobots and AI has huge potential to meet these requirements.

A cobot is a general-purpose humanoid robot that can automate within a human-centric production line. A cobot-centric platform doesn’t require retooling for conveyer belt assembly lines staffed by huge single-arm gantry-like devices that only do one function. Instead, the cobot is placed in the human-operated space and can perform the same physical interactions as a human being. Now, automation can work in existing production lines — a critical need for aerospace manufacturers. Furthermore, a single cobot can perform multiple tasks. There is no need to design physical automation task-by-task or function-by-function. This is a critical factor for bringing automation into the industry because cobots can quickly drive return on investment by performing multiple functions in a hand-driven, low-volume plant.

Software guides the cobot

The physical aspect of automation is only half of the equation. Software is needed to guide and drive the automated systems, and that’s a significant cost driver in adopting automation. In the automotive industry, such software is procedurally written, which makes sense for performing only a single function. Writing procedural software to guide a cobot across a wide range of human-like work in a human-driven production line, on the other hand, is a daunting if not impossible task. This is where AI comes in.

The fast advancement of complex operational AI — think autonomous driving — has led to the creation of robust development tool chains and methodologies. These can be adapted for different use cases, such as factory automation. By automating the data gathering, curation and processing across a complex-stage AI development program, the cost of developing the automation software shrinks significantly, as it essentially becomes an automated function itself.

Cobots+AI will not give the industry a turnkey solution to factory automation on day one. But this new option will eliminate the cost-prohibitive barriers to adoption. With the right AI development toolchain in place, aerospace manufacturers can introduce automation successfully at low cost, and in a fraction of the time it would take to design traditional automation. Now they can be ready to supply the air taxi segment at high volume.