A single compromised password should not be enough to expose your drawings, stop an ERP system or give an outside supplier access to a production network. That is the practical case for zero trust in manufacturing. It is an approach that checks every request for access rather than assuming that a person or device is safe simply because it is inside the company network.
For a manufacturer, this is not about making staff jump through needless security hoops. It is about limiting the damage when an account, laptop or supplier connection is compromised, while keeping production moving. Applied properly, zero trust can protect the systems that matter most without forcing a risky, disruptive replacement of every older machine on site.
Zero trust starts with a simple change in assumption
Traditional networks often work on a trusted-inside, untrusted-outside model. Once a user signs in at the office, connects through a VPN or plugs into the factory network, they may be able to reach far more than their role requires. This made sense when systems sat in one building and few people worked remotely. It is a poor fit for a modern engineering or manufacturing business with cloud applications, mobile devices, shared terminals and remote support arrangements.
Zero trust replaces that broad assumption with a more useful question: should this user, on this device, be allowed to access this specific system right now?
The answer depends on evidence. The user may need to prove their identity with multi-factor authentication. Their device should be known, encrypted, patched and protected. Their role should permit the requested task. The application or file should be available only from an appropriate network location. If something changes, such as a sign-in from an unfamiliar country or an unpatched laptop, access can be blocked or challenged.
This does not mean trusting nothing and nobody in a literal sense. It means granting limited, verified access rather than handing out permanent keys to the whole site.
Why zero trust matters on the factory floor
Manufacturers tend to have more dependencies than a typical office. A production planner may rely on an MRP system, a shared file store and a label printer. A machine operator may use a computer connected to a CNC machine that cannot tolerate unplanned changes. An engineer may need CAD files stored in Microsoft 365, while an external specialist needs occasional access to diagnose a fault.
If all these systems share a flat network, a phishing attack on an office account can become a much bigger problem. An attacker may move from one computer to another, find shared folders, reach backups or interfere with systems that support production. The first affected mailbox may be a nuisance. The missed delivery and recovery work are where costs start to build.
Zero trust reduces the routes available to an attacker. It separates office systems from machinery-connected devices, restricts access to sensitive design data and gives suppliers a narrow, controlled route to the equipment they support. It also makes it easier to see who accessed what, which matters when a customer asks sensible questions about how intellectual property and production information are protected.
Zero trust is not one product or a switch to flick
Some suppliers present zero trust as a product purchase. In reality, it is a set of decisions across identity, devices, networks, applications and data. Microsoft 365 security settings may form part of it. So might managed devices, network segmentation, secure remote access and better logging. The right starting point depends on your current risks and operational constraints.
For a 20-person precision engineering firm, the immediate priority may be multi-factor authentication, removal of shared administrator accounts and protection for managed laptops. For a larger site with several production cells, separating the operational technology network from the office network may be more urgent. Both are valid zero trust improvements.
Trying to deliver every element at once can create unnecessary disruption. A staged plan is usually safer. Start with systems that could stop production or expose valuable information, then improve the supporting controls around them.
The practical building blocks of zero trust
Verify people before granting access
Every person should have an individual account. Shared logins are common around workshops and warehouses, but they remove accountability and make it difficult to revoke access when someone changes role or leaves. Where a shared terminal is necessary, use separate named sign-ins where the software allows it, or put compensating controls around the device.
Multi-factor authentication should protect email, cloud services, remote access and administrator accounts as a priority. It is not flawless, and users can still be tricked into approving a prompt, but it greatly reduces the value of a stolen password. Number matching, conditional access rules and user awareness training help reduce that risk.
Permissions should follow job roles. A buyer does not need access to CAD folders. A production operator does not need administrator rights. An IT provider does not need unrestricted, permanent access to every server simply because it is convenient.
Check the health of devices
A laptop that handles drawings, quotations or production planning should be managed, encrypted and kept up to date. If it is lost, the business should be able to remove company data or prevent it being used to access systems. This is especially relevant for directors, engineers and sales staff who travel between sites and customers.
Not every device can meet the same standard. A Windows 7 computer controlling a machine may be unsupported, and updating it may invalidate the machine supplier’s support arrangement. Pretending it is a normal office PC is how problems get created.
Instead, identify the constraint and reduce exposure. Put the machine computer on a separate network segment, block its internet access unless it is genuinely needed, allow only specified communication paths and maintain a tested backup or image of its configuration. Document who can access it and how support is provided. That is zero trust applied with operational common sense.
Segment networks around production needs
Network segmentation is one of the most valuable controls in a manufacturing environment. It creates boundaries between areas such as office devices, guest Wi-Fi, warehouse scanners, servers, backup systems and machinery-connected equipment.
The objective is not to stop useful communication. Your ERP system may need to exchange data with a shop-floor terminal, for example. The objective is to permit that known connection while blocking everything else by default. This limits lateral movement if one area is compromised and makes faults easier to investigate.
Segmentation needs careful planning. Poorly designed rules can stop label printing, machine monitoring or data collection at the worst possible moment. Map the dependencies first, test changes in a controlled window and keep a clear rollback plan. Security that halts a shift has missed the point.
Control supplier and remote support access
Remote supplier access deserves particular attention. A machinery supplier may need rapid access when a fault threatens an urgent order, but an always-on connection to the production network is a significant risk.
A better arrangement gives approved suppliers access only when needed, using named accounts, multi-factor authentication and a controlled remote support method. Access should be restricted to the relevant equipment, time-limited where possible and reviewed regularly. Keep records of who has access, including former suppliers and contractors. Old accounts have a habit of surviving longer than the equipment they were created for.
Protect data and recoverability
Zero trust also applies to data. Restrict sensitive folders, design files and commercial documents to the people who need them. Apply sensible controls to prevent accidental external sharing. Retain audit records for important activity where your systems support them.
Backups remain essential, but they are not a substitute for access control. A ransomware attack can still cause days of disruption even when data is recoverable. Backups should be separate from normal user access, monitored and tested against the applications that matter, including ERP or MRP databases and machine configuration files. A successful restore of a few documents is not proof that production can resume.
A sensible zero trust plan for a manufacturer
Start by identifying the systems whose failure would stop output, delay despatch or expose valuable information. Include production PCs, ERP and MRP systems, CAD and CAM data, email, file storage, backups, Wi-Fi and remote access. Then establish who uses each system, which devices connect to it and which suppliers need access.
From there, address obvious weaknesses first: unsupported remote access, shared privileged accounts, missing multi-factor authentication, unmanaged laptops and unrestricted links between office and production networks. Record exceptions, particularly legacy equipment, with an owner and a plan to reduce risk over time.
This work can support customer assurance requests and Cyber Essentials preparation, but good security practice and formal certification are not the same thing. Certification scope and requirements need to be assessed against the current scheme rules and your own environment. A network workaround may reduce risk without automatically making a system compliant.
For businesses in Hampshire, Surrey and West Sussex, the most useful next step is usually a practical review of access, devices and network boundaries around critical production systems. Syn-Star can help you turn that review into a phased plan that protects uptime first, then steadily reduces the risks that could put it at stake.
