A Remote Operations Service for Keeping Stalled Smart Factories Running

Installation is finished, but operations remain

The Ministry of SMEs and Startups, Korea’s government ministry responsible for small and medium-sized enterprises, directly surveyed 5,000 manufacturing SMEs and mid-sized companies from a base of 163,273 factories. Smart-factory adoption was 19.5%, or roughly one in five factories, and 75.5% of adopters remained at the basic stage.

An analysis of the underlying data by the Korea SMEs and Startups Institute found that 92.4% of adopting companies collect production data. But 75.7%, more than seven in ten, entered it manually, while only 19.7% collected manufacturing data in real time.

Among companies that analyze the data they collect, 76.0% used general document tools such as Excel. That means the work of downloading files from a production programme, pasting them into Excel, and calling the shop floor when the figures do not match still remains.

The operating foundation is weaker still. At 99.2% of companies, there was no organization responsible for manufacturing data and artificial intelligence, and only 6.6% set aside a separate budget for smart-factory training. Even when a system works, there are very few people available to teach others or trace errors.

The Korean government also supports parts repairs, programme improvements, and responses to process changes through its 2026 aftercare programme. It provides up to KRW 19 million in government funding per company to around 210 companies nationwide, with the company paying half of the cost. That leaves separate room for a private service that continually handles small, everyday errors.

A place to turn three rounds of transcription into one

The following scenario is an assumption intended to make the survey figures easier to understand. Imagine a factory with 24 employees making plastic parts, where one production-management team leader also runs the smart factory.

The team leader writes customer orders into an Excel production plan, then enters the same information again into a production-management programme. When a shop-floor supervisor records production quantities on paper or on a work screen, the team leader downloads those figures in the evening and pastes them again into the weekly-report Excel file. The same quantity passes through three places.

If a value is missing from a sensor that reads equipment status, or if an item number is entered differently, the figures on the production screen and in Excel no longer match. The team leader takes a screenshot of the error and calls both the installation company and the equipment supplier, explaining the issue from the beginning to each one.

What is needed here is not another production programme, but a single error-intake window. The team leader uploads the error screen, time of occurrence, and equipment name. With permission, an external operator reviews the operating logs, then separates the case into an input mistake, a programme setting issue, a sensor connection issue, or a fault requiring an on-site visit.

The intake window also records the responsible supplier, the next action, and the expected response time. Work that had been scattered across phone calls, the production programme, email, and Excel becomes one record, and the team leader does not need to repeat the same explanation.

A weekly report can also begin as a draft built from existing data. Items where production quantity and defect quantity do not match should be flagged rather than changed arbitrarily. The shop-floor supervisor can record the reason, and the team leader can make the final approval.

Some work must clearly remain with people. Actual equipment repairs, safety checks, and final judgment on incorrectly entered numbers belong to on-site staff. A remote operator should focus on classifying issues and connecting records.

A suitable first customer is a factory with 10 to 50 employees making food products, metal-processed goods, or plastic parts. It already has a production-management programme or sensors but has no dedicated operations staff. Rather than taking responsibility for the whole factory from the start, begin with one production line. Accept only quantity mismatches, screen errors, and missing reports, then record resolution time and whether each issue repeats.

Other countries are selling operations separately

Japan paired a small monitoring screen with an improvement map

Japan’s ISTC provided small and medium-sized manufacturers with a service that automatically collected and displayed production quantities and downtime from existing equipment. A simple conversion of its publicly listed price at the time puts the initial cost at about KRW 850,000 and the monthly cost for five production lines at about KRW 340,000. Its current price needs to be checked again.

The company did not provide only a screen. It added a diagnostic report that read five days of accumulated data and explained problems and improvement methods. In a case presented by the supplier, a customer factory used its own analysis and improvements to raise hourly output by 59% over three months. This was a company-published figure and should be treated accordingly.

Canada separated causes remotely before dispatching a technician

Canadian equipment manufacturer Conceptromec installed remote-access devices at customer factories so its head-office technicians could review equipment-control screens and production status. When a fault was reported, it checked the cause remotely first instead of immediately sending a technician.

Device supplier HMS Networks said it resolved 90% of service requests remotely from its head office. The claim requires independent verification because it comes from the supplier’s own case presentation. Still, it shows that rapid cause classification and remote recovery, not just new equipment, can be something customers pay for.

Germany organized existing programmes instead of buying a new one

Metal furniture manufacturer Metallmanufaktur Rammro received support to organize work instructions and manufacturing knowledge scattered across paper, Word, and Excel. Expert support, demonstrations, and advisory services from a regional digital support center were free. The company paid about KRW 600,000 for workshop costs charged by its existing enterprise-management programme supplier.

The result was not a new programme purchase. The company used its existing enterprise-management programme to organize manufacturing knowledge into PDFs for shop-floor staff to view online, and it avoided buying additional user licences. This is an example of fixing disconnections between tools already in use as a distinct service.

Things you could build from this now

1. A remote factory error inbox

• What it does: A service that receives screen errors and equipment issues in one place, then routes them to remote resolution, supplier handoff, or an on-site visit.
• Who uses it: A production manager at a 10-to-30-person plastic injection-molding factory after the installer’s free repair period has ended.
• Why now: At factories with almost no separate operations organization, the person in charge is repeatedly explaining the same symptom to multiple suppliers.
• First screen: A large intake button where a factory employee enters only a photo, the time of occurrence, the equipment name, and whether production has stopped.

2. A weekly check for number mismatches

• What it does: A service that compares quantities in the production programme and Excel reports, then flags missing entries, duplicates, and item-number mismatches each week.
• Who uses it: A 20-to-50-person food-processing factory where workers manually enter production volumes and temperature records.
• Why now: Many companies collect data, but more than seven in ten enter it manually, creating substantial room for errors.
• First screen: Show this week’s mismatch count, the source number, the report number, and the person who needs to check each item side by side.

3. Ten-minute training built from fixed errors

• What it does: Turns one resolved error into a photo and short explanation that shop-floor staff can revisit on their phones.
• Who uses it: A 15-to-40-person metal-processing factory where a supervisor personally teaches new workers how to use the production-management screen.
• Why now: Fewer than seven in 100 companies set aside a separate training budget, making short on-site training more realistic than long external courses.
• First screen: Show one error to learn today, the correct input sequence, and one final check question.

4. A smart-factory handover ledger

• What it does: Organizes each factory’s installer contacts, repair periods, programme accounts, data-storage locations, and recent changes.
• Who uses it: A 10-to-50-person manufacturer that built a system through a support programme, but whose original contact has left or whose supplier contact point has changed.
• Why now: Before diagnosing a fault, people often spend time finding who to contact and what contract remains.
• First screen: Use traffic-light colors to show the factory’s system list, responsible supplier, repair period, and date of the last data backup.

Why this matters where you are

Korea’s figures describe a specific local condition: government-supported smart-factory adoption, limited internal operating teams, and an aftercare programme that requires companies to share costs. In your own market, check whether factories also rely on manual data entry, scattered supplier contacts, and existing tools that do not connect. If they do, test a narrow service around one production line before proposing a replacement system.

What to check today

Ask a production manager at one factory using a smart-factory system for 30 minutes. Have them draw, in sequence, the screens, Excel files, and phone contacts involved in resolving one recent error. If one case moves across three or more tools and the responsible supplier cannot be identified within 10 minutes, it may be worth testing a remote intake service for one production line.