Who let the robodogs out? TCS awaits clearance for physical AI in the aerospace sector.

The aerospace industry is in transition as it prepares to develop greener aircraft by adopting new materials and propulsion systems, to integrate digital technologies such as physical AI into its processes and to expand into new markets such as urban air mobility. 

For TCS, aerospace is one of the biggest sub-sectors it works with in its manufacturing service line. It provides services to several large plane manufacturers as well as significant players in the aerospace supply chain. According to Anupam Singhal, President and Business Group Head of Manufacturing for TCS, it is an extremely important sector for TCS’s growth. 

Re-imagining the aerospace industry

TCS surveyed 323 senior aerospace leaders across North America and Europe to provide a snapshot of their main concerns and plans. The survey found that respondents anticipate 40% of their manufacturing operations will be lights-out within the next five to seven years. Ambitions for the adoption of robotics within the sector remain high but the ability for such a highly regulated industry to run “lights-out” operations is unlikely. This means that human-led maintenance will remain the norm in this sector for the foreseeable future.

Use of digital twins is rising across the aerospace ecosystem and is seen by survey respondents as the most impactful technology on the sector by 2035. Digital twins are enabled by digital threads which are the connected data and workflows that span the lifecycle of the product being modelled. Consequently, the fact that the survey found that 60% of respondents have begun implementing digital threads suggests that executives are maturing in their use of digital technologies and adoption of digital twins.

There has been talk of commercial electric vertical take-off and landing (eVTOL) vehicles for some time, so the survey finding that nearly three quarters of Advanced Air Mobility (AAM) companies are actively building or involved in commercial projects indicates committed investment in the area. Nevertheless, 80% continue to cite regulatory uncertainty as the biggest challenge for the development of this market in waiting, where for example air taxi fleets in areas such as New York are being developed and work is ongoing with air traffic control to enable vertical take-off and landing.

Regulation and innovation

Singhal says in his view the aerospace industry is about in the middle of the manufacturing sector pack for adoption of digital technologies, with the chemical industry and process manufacturing areas ahead in use of digital twins and digital threads.He adds:

We used our first digital twin back in 2011 and saw it as a real enabler but the adoption path has been slow, yet it is now coming into its own again aided by AI and significant increase in compute power so that the value realised and the acceptance is now much higher. For example, airlines only make money when the plane is in the air so it is important to use a digital twin of the engine in the air. We have spent over a decade working with Indian manufacturers so that they can do predictive maintenance which all feeds into the customer experience as you don’t then face delays because of mechanical issues.

This industry requires a lot of innovation and it is commercially possible to do. Technology can be a huge aid to help scale innovation quickly at a cheaper cost especially when it comes to the next generation of planes as we rely more on digital engineering than on mechanical engineering. 

Robots

In this vein TCS has developed physical-AI powered quadruped robots that are equipped with depth cameras, LiDAR, environmental sensors, and on-board edge processing units to drive operational intelligence. As Singhal explains:

It can walk around in areas where humans cannot and identify problems using sensors and can take corrective action.

A case in point is the thousands of rivets on an aircraft fuselage where there is a large opportunity for physical AI to execute repairs and inspect the state of the aircraft.

Although not used in the aerospace sector yet, TCS has used its quadruped to set up a new semiconductor plant for a client. Singhal explains:

It can be used to ensure everything is being done properly and if it is not, it lets the plant know as it can capture the smallest defect. It is an example of where AI on the edge is important. It can do final inspection much quicker than human intelligence can.

He adds:

Physical AI could do a lot more but it is difficult to introduce in highly regulated processes. It can make businesses more productive by using all the data that needs to be analysed across segments and industries to improve productivity and quality in order to deliver superior customer experience.

My take

The aerospace industry has full order books for the next decade so there is a significant amount of opportunity there for service providers such as TCS to help the industry increase throughput and better service the demand backlog. While the TCS research indicates that technology appetite is outpacing regulatory approval, the increased adoption of AI on the edge to improve quality and time to market are growing. 

Although TCS does not work directly with regulators, it is present at aerospace events where it showcases capabilities such as its quadruped to show how much quicker the robodog can undertake, say, a reliable landing gear inspection. OEMs are beginning to see the attraction of physical AI, and regulators are more disposed to approve use in hazardous environments, so the industry is hoping some aerospace use cases may get the go ahead in the near future.