Korean drugmakers are commissioning facilities designed to pass U.S. and European inspections on the first attempt, and they want those requirements fixed in the blueprint before concrete is poured.

AtkinsRéalis, the Montreal-based design and project firm that entered Korea in 2007 and opened a Seoul office in 2009, has built its growth here on that premise.

Korea designs for first-pass FDA and EMA signoffs

“We’re the only international engineering firm in Korea with a full local pharma team,” said Kim Jin-hong, Country Head and Director of Industrial South Korea at AtkinsRéalis, a post he has held since 2012.

He said the office has about 120 engineers, including 20 to 25 in-house process designers, along with architects, piping, automation and project delivery staff. “If we do the heavy discussion at the start, we can prevent the design changes and delays that appear later.”

The company rebranded to AtkinsRéalis in 2023 and has standardized a digital method that begins in basic design. “From basic design, architecture, structure, mechanical, electrical, process and piping all design in 3D at the same time, in the cloud,” Kim said.

The live model is then linked to construction sequence and cost, so design clashes and installation conflicts are detected in the model and resolved before procurement and field work.

New work in Korea reflects that discipline. In July 2024, AtkinsRéalis won process detail design, automation and procurement support for Merck Life Science’s bioprocessing production center, with Kolon Global Corporation as the engineering, procurement and construction (EPC) partner delivering the facility.

Plans call for advanced production units, an automated warehouse, a distribution center and on-site quality-control lab services across roughly 43,000 square meters.

Around the same time, the firm started basic-through-detailed multidisciplinary design for Hana Pharm’s new injectables facility in Pyeongtaek, where an environmental-management backbone and a scalable automated warehouse are included from the outset to support future expansion and logistics.

Owners, Kim said, still run into the same traps: compressing room sizes to meet near-term budgets and postponing equipment choices that later force expensive layout changes. 

Korean sponsors often approach concept design as a floor-plan exercise, focusing on “perfect flows” on paper for personnel, product, material and waste, while underestimating the GMP requirement for working space and service access. “GMP’s first requirement is sufficient space,” he said. “If you delay that decision, the cost of late changes is enormous.”

He described Korea as an attractive place to execute, with more predictable schedules and a stronger supplier base around major hubs than five years ago. The pressure points are known: long lead times for specialized equipment, a limited pool of subcontractors with deep GMP experience, and governance models that route large builds through group-affiliated construction arms. 

Rather than unwind those structures, the firm argues for inserting specialist design teams early and putting a single, live model in front of owners, general contractors and niche trades so decisions about flexibility, redundancy and validation burden are made against the same geometry, sequence and cost data.

Where capital is heading and what it takes to get there

Repeatability is where schedules compress. Kim described a long-running blood-plasma program but declined to identify the sponsor on the record. “We designed the first Andong plant in 2014, replicated the process in Indonesia and now the same team is moving to Europe,” he said. 

AtkinsRéalis led the Andong design and worked with its Southeast Asia offices on the Indonesian build, now under construction; a European site is being prepared. 

When a process, its documentation and work spaces are already proven, Kim said, commissioning and qualification shorten because teams reuse validated designs and procedures rather than rebuild them.

AtkinsRéalis considered Japan and China when it chose an Asia base years ago, Kim said, but settled on Seoul because foreign design teams could carry a project further under local rules and because policy support and local capability were rising.  

He remains comfortable with that choice. He described China as a market of outsized scale with mandated handoffs to designated local design institutes at detailed design, and Japan as a large but difficult market to enter. 

Korea, by contrast, has what he called an end-to-end execution culture and a life-sciences ecosystem that continues to mature.

Capital, he said, is moving toward sterile injectables and specialty lines with tighter containment and heavier automation. “ADC is very hot,” he said, referring to antibody-drug conjugates that require both biologics production and high-potency small-molecule handling. 

He also sees rising interest in controlled-release platforms, including microsphere technologies that lengthen drug residence time and change facility logic by shifting where and how materials are handled. Classic biologics capacity is still being added, he said, but can tip into overbuild when sponsors assume tech transfer and long-lead utilities will sort themselves out late in the schedule.

The weak link he encounters most often is organizational. Large groups in Korea tend to route projects through affiliated builders that may lack deep GMP backgrounds, which can erase the benefit of aggressive timelines if specialist design is not in place. “Don’t approach this like building an apartment,” Kim said. “Bring in a specialist designer from the start.” 

Smaller R&D-heavy companies, by contrast, often lack internal teams and hand end-to-end project management to outside engineers; that can work, he said, if requirements are frozen early and equipment orders go in before procurement windows close.

Digitalization is the near-term lever. AtkinsRéalis now starts most Korea projects in full-discipline 3D at basic design, shares models in the cloud and overlays construction sequence and cost as scope hardens, which reduces rework before steel goes up and shortens design cycles. 

The longer-term target is a true digital twin in operations, where live data from utilities, environmental monitoring and equipment performance feed models that optimize campaigns and maintenance.

That step, Kim cautioned, requires owners to fund the data spine at design. “In facilities and operations, there is not yet a real project in Korea that applies AI as a smart factory or digital twin,” he said. “That demand will come.”

Not every technology push turns into local factories. Venture-stage programs often sell intellectual property near commercialization rather than invest in full manufacturing build-outs, which reduces the number of R&D stories that become plants. “Many ventures sell the technology before commercialization,” he said. “It’s understandable, but still unfortunate.” 

He argued that Korea could capture more downstream value if sponsors partner earlier with manufacturers and fund complete production trains from the start: rooms sized for people to work, equipment specified as purchased, and utilities and monitoring designed with inspection in mind.

The company has been cautious about public naming, limiting references to already disclosed projects. It does not present itself as a one-stop builder. What it offers, Kim said, is a single source of geometry, sequence and budget that treats inspection readiness as a design input rather than a last-mile retrofit. 

Asked what would move Korea up the regional ranking on speed, predictability and total delivered cost, he did not reach for a policy rewrite. “Size critical rooms so people can do the work, and order long-lead equipment early,” he said.

The rest, he added, follows from those choices: utilities specified to the actual load, monitoring and data integrity built into the rooms that will be validated, and documentation that matches what is installed.