Metal fabrication is much more than using steel sheets and welding techniques to construct a structure or weldment. Though taking genesis there, the progression of a metal fabricator is typically rapid to value-added activities that involve the addition of multiple components to a basic fabrication and transforming it into a functional machine. These, without imposing limitations might include electric motors, bearings, drive parts, pneumatic and hydraulic fittings, electrical hardware and machine programming to make them all work in sequence. Therefore, it is beneficial to look at the impact of a disruptive concept such as AI through the eyes of a machine builder.
Machine builders often start with a concept and thrive in the pride of having ‘sold that concept’ to the end-user and supply a machine with moving parts. Over time, this machine builder’s curated concept, with the right traction evolves into a standard product with its unique life cycle. This transformation travels through the painful journey of cost-overruns, design flaws leading to redesigns, warranty costs and upgrades. Starting with reasonable estimates, tracking costs and utilizing this valuable data to direct future outcomes are key to the machine builder’s sustainability. In other words, Predictability is a key feature of AI that holds promise to help this machine builder. Let us analyze the different stages of this value chain:
Cost estimates
Mathematical modeling can use data from past projects (costing versus actual spend), current market requirements (efficacy of present product features and new customer demands such as material strength, energy efficiency etc.), new developments (e.g. Industry 4.0 capability and reporting), supplier inputs (tariffs, freight increases) to arrive at close to accurate cost for a new product.
In addition to helping the machine builder with appropriate pricing, the reach of such models can be extended to schedule manufacturing activities, source key inputs at a time that is conducive to eliminate holding high value inventory. Sourcing leads to sell this product to can be very effective with the right CRM software, something that prudent machine builders are actively engaged in.
Machine functionality
- This is an aspect where AI can play a major role in several areas. In so much as machine breakdowns and maintenance proper use of data can accurately predict failure modes and prepare the user to stock and replace wear parts before actual wear takes place.
- Machines that are part of a production cell could be dictated by other machines in the same cell, typically the one with the longest cycle time. Breakdown predictability is very critical in this production environment since the entire process could come to a grinding halt with the failure of a single component in one of the cell members.
- Data obtained from multiple sensors in the machine unless analyzed and utilized to improve the machine performance, tends to be a wasteful collection of useful information. Most PLCs in modern machines are very capable of collecting such information but tend to be grossly underutilized much like a common calculator or mobile phone!
- Data collected when the machine is in production at the end-user’s plant is the most valuable data in terms of efficacy. When permitted, this data can be used to target the customer with an offer for spare parts and maintenance. In most situations, the customer is only too happy with the knowledge that their machinery is being monitored, and the onus of spare parts replacement has been outsourced to the OEM.
Observation and conclusion
AI is looked at through two different lenses whether you are that business owner with a penchant for growth or the anxious employee that views it with the same trepidation as automation – both leading to job losses. Though there might be some truth to this statement, automation resulted in not just re-tooling of the process, but also the operators. Those that chose not to accept the challenge had to contend with a lower-skill occupation with corresponding remuneration.
Metal fabrication is a large part of machine building. This industry has already seen digitization of their production process (digital routers to replace technique sheets, job costing, supply chain management have been digitized for several years). Infusing AI into this factory automation will give it the steroidal treatment that businesses are looking for.
This discussion cannot be complete without acknowledging what all this means to the consumer. Would these ‘smart’ initiatives lead to cost-savings at the pedestrian level? Will operating efficiencies increase to the extent that costs of regularly used products such as cars will see a drop? Would employee training be a critical part of operating efficiency? With the amount of data being drawn from existing machine controls, does this expose a security threat, especially with sensitive information. This transformation of manufacturing including metal fabrication is not reversible. Cost savings are critical for most in this sector to be sustainable. Rejection of AI is a greater threat than the potential ills of adopting it.
