3T-AM, a Newbury-based metal additive manufacturer, offers a concrete case study in what changes when a prototype shop tries to become a production supplier. Commercial director Eddie Andrews, who joined in November 2021, frames the problem in manufacturing terms: the company already knew how to make a one-off component for hundreds of customers, but it had to learn how to win the same order again and deliver it at the agreed price and schedule. Five years later, repeat orders account for around 70% of revenue, while new product introduction still consumes roughly 80% of the team’s time.
A different order book
Andrews’s distinction between revenue and time is central. Production work pays the bills, but development work keeps the pipeline alive. Semiconductor customers generate the strongest demand, followed by energy, then space and defence. Aerospace remains central to the company’s plans, but 3T-AM has found work for its machines while it waits for more aerospace programmes to reach suppliers at volume. The company dates to 1997 and its first metal machine arrived around 2008. By the 2021 reset, years of accommodating individual prototype jobs had left a shop floor with process after process. 3T-AM narrowed its customer base, gave larger accounts more influence over development and simplified accumulated procedures. Much of the work remained manual while the team established which operations it could reliably repeat.
Who owns the finished part?
Changing who owned a job mattered as much as changing equipment. When the first person to see a drawing was the engineer responsible for printing it, the natural question was how to print it, and decisions at that point shaped work handed to everyone else. Today a product owner is responsible for delivering the finished component, and print and machining specialists agree on the condition of the part as it passes between operations. Andrews’s formulation is blunt: to become a successful additive manufacturer, a company needs to be a really good CNC machining company. Bringing machining in-house gave the SME more control over costs, quality and delivery dates. Larger multi-laser machines improved productivity at the printing stage, but neither investment on its own could tell a customer when a finished part would arrive.
Buying machines against demand
Business model and capital allocation now reinforce each other. Previously, 3T-AM might have bought an EOS M400 to try a different material and see what business it could develop. Today, according to Andrews, customers’ programmes account for the capacity of incoming machines before they reach the factory. The investment, he says, is quite boring. Austrian precision manufacturing group SBO acquired 3T-AM in October 2025. The Newbury business had generated around £5 million in sales in 2024. SBO brought experience in machining downhole drilling equipment, including components up to ten metres long. What interested the group was 3T-AM’s metal printing expertise and access to a different set of industrial customers. SBO approved a five-year business plan and asked whether it could be done quicker. In August 2026, the group reported that 3T-AM had added an EOS M400 in the second quarter, ordered another printer and invested further in post-processing as part of a seven-printer expansion across UK and US operations.
A full order book makes the next machine easier to justify, but it also leaves little time on that machine to qualify a different material or develop a new product. Andrews wants to preserve two distinct reasons to invest: contracted demand and the work 3T-AM cannot yet offer. A pre-pandemic aerospace enquiry illustrates the scale of a production commitment. A single family of parts would have required 16 multi-laser systems to cover only a share of the requirement. From the buyer’s side, the question was how many suppliers had the floor space, the capital and the capacity to take it on. Andrews sees that as one reason original equipment manufacturers still keep sizeable additive programmes in-house.
What the customer can measure
For some customers, replacing a casting with a printed part earns its keep through lead time alone. A legacy or development programme can obtain a component sooner, even at much the same unit cost. At higher volumes the calculation gets tougher. Machine maintenance contracts and consumables weigh on the hourly rate, and a factory working a single shift has fewer productive hours over which to spread those costs. Andrews sees scope to learn from CNC shops that operate around the clock. One place he would look for automation is support removal, where the process can still depend on what an experienced operator knows about a particular job.
A project with Honeywell Aerospace will put another claim for additive manufacturing to the test. Under an Aerospace Technology Institute-funded programme, the partners are converting an existing conventionally manufactured component to additive. The sustainability team at the NCC in Bristol is carrying out a 12-month life-cycle assessment covering manufacture, service and end of life. Printing closer to the finished shape should use less feedstock and leave less metal to machine away than starting with billet. The study will compare the two manufacturing routes alongside any benefit the redesigned component delivers in service. The intended outcome in Newbury is something a purchasing department could use. As 3T-AM connects its factory and enterprise systems, it is developing digital product passports to document the carbon associated with making an individual component. By the end of 2028, Andrews wants that information to come with every part at no extra charge. For a customer compiling a supply-chain emissions report, it would become one more thing they could expect from a routine order.
The boundaries of the model
3T-AM’s experience does not describe a simple switch from prototypes to production. The commercial transition was harder to time. Production orders rose and then tapered off. Turning away too much prototype work during the upswing would have left gaps when it eased. Repeat orders now give 3T-AM room to choose which development jobs to take, but some will never become production contracts. Andrews’s argument is that participation is still necessary: if you do not take part in the journey, you are never going to find out. That is a useful corrective to the idea that production work alone can carry an additive manufacturer. The model depends on a mixed order book, a product owner who can arbitrate between printing and machining, and cost calculations that customers can see. It also depends on capacity that is committed before investment, rather than capacity built on the hope of finding work.
The more demanding test is whether production discipline can coexist with the development work that creates future programmes. A full order book may make capital spending easier, but it also reduces the slack needed to qualify materials and prove new designs. 3T-AM’s answer is to keep both reasons for investment visible and to turn the resulting data into a routine part of the order. Whether that becomes a durable advantage will depend on the same things Andrews describes as boring: repeatable processes, agreed hand-offs, realistic hourly rates and delivery dates that a customer can plan around.
