AI Is Moving from the Screen to the Production Line
You may think of artificial intelligence as a tool for writing marketing copy, answering customer questions, or summarising reports. But the more important shift may be happening behind the scenes: AI is increasingly helping businesses design, inspect, and improve physical products.
That is the lesson behind Apple’s new foldable phone, the iPhone Duo. According to TechCrunch, Apple used AI algorithms to match each hinge with its best-fit housing. It also used laser scanning and 3D printing to correct tiny surface imperfections during manufacturing.
For a Malaysian SME, this is not only a story about an expensive smartphone. It shows how automation can support quality control, reduce inconsistency, and help a small team handle more detailed work without relying entirely on manual inspection.
TL;DR
Apple’s foldable-phone hinge shows how AI can be used to improve physical production, not just office tasks.
You can apply the same principle in your SME by using digital inspections, automated matching, and production data to catch defects earlier and make quality more consistent.
What This Means
A foldable phone hinge must open and close repeatedly while keeping two panels aligned. Small variations in the hinge, housing, or surface can affect how smoothly the phone operates and how long it lasts. Apple says its manufacturing process uses AI to match individual hinges to suitable housings, rather than treating every component as exactly identical.
The process also includes a confocal laser that scans the surface of each unit. Where the scan identifies residual waviness, the company uses 3D printing to add as many as 25 micro layers of a custom photopolymer to correct the surface. These details were reported by TechCrunch.
In plain language, the factory is not simply assembling parts and checking a sample at the end. It is measuring each unit, making decisions based on the measurements, and applying a correction where needed. That is a very practical form of AI: observe, compare, decide, and improve.
The useful question is not “Where can you add AI?” It is “Where does your business make the same quality decision repeatedly?”
This distinction matters. Many SMEs start with general-purpose tools because they are easy to try. However, the biggest operational benefit often comes from a narrow workflow where your staff repeatedly inspect items, match information, identify exceptions, or correct errors.
How This Applies to Malaysian SMEs
If you run a food manufacturing or catering business, consistency is a daily challenge. You may need to check portion sizes, packaging labels, expiry dates, temperatures, or the appearance of finished products. A simple digital inspection process can let staff record results using a phone or tablet, while automated rules flag readings outside your accepted range. You do not need a sophisticated factory robot to benefit. A structured checklist, photo evidence, and automatic alerts can already reduce the risk of missed checks.
For workshops, fabricators, and small manufacturers in places such as Shah Alam, Johor Bahru, Penang, or Klang, the same idea can apply to component matching and defect detection. Suppose different parts have slightly different tolerances. Instead of asking a worker to remember which component belongs with which batch, your system can compare measurements, lot numbers, or job specifications and recommend the correct pairing. Staff still make the final decision where necessary, but the repetitive comparison is handled consistently.
Retailers and distributors can also use this approach in inventory operations. A warehouse worker may receive goods, scan product codes, compare quantities against a purchase order, and take photos of damaged cartons. Automation can identify discrepancies and route exceptions to a supervisor. This is similar in principle to Apple’s hinge process: inspect every relevant unit, compare it with the expected standard, and focus human attention on items that need judgement.
Service businesses have an equivalent opportunity. An air-conditioning contractor, for example, can use a mobile form to record equipment condition, pressure readings, photos, parts used, and follow-up requirements. The system can flag incomplete records or unusual readings before the technician closes the job. A cleaning company can require time-stamped checklist completion and photo verification for important areas. These workflows create a repeatable quality standard even when different employees handle the work.
The key is to begin with a clearly defined process. AI cannot compensate for an unclear standard. First decide what “acceptable” means, what information must be captured, and which exceptions require a manager’s attention. Then automation can help apply that standard more reliably.
A Simple Comparison
| Manual approach | Automated approach | Useful SME outcome |
|---|---|---|
| Inspect a sample after production | Record checks for every relevant unit | Earlier detection of recurring defects |
| Match parts using memory or paper notes | Compare item codes, measurements, or batches | Fewer assembly and fulfilment errors |
| Review all records manually | Flag missing or unusual entries | Managers focus on exceptions |
| Correct problems after delivery | Identify patterns during the process | Better prevention and follow-up |
Apple’s reported process includes up to 25 micro layers of printed correction material on a unit. That number is specific to Apple’s manufacturing method and should not be copied directly into your operation. The broader lesson is more useful: precise measurement can enable a targeted correction instead of a full rework.
Practical Takeaways for Your Business
- Choose one repeatable quality decision. Look for a task your staff perform every day, such as checking dimensions, matching orders, reviewing photos, or validating forms.
- Define the acceptable range. Write down the standards, limits, and conditions that separate an acceptable item from an exception.
- Capture evidence at the point of work. Use photographs, barcode scans, measurements, timestamps, or digital signatures where appropriate.
- Automate alerts, not every judgement. Let the system identify unusual cases while experienced staff handle decisions that require context.
- Track recurring exceptions. If the same issue appears repeatedly, investigate the process, supplier, training, or equipment behind it.
- Keep a human approval step for important decisions. Automation should support accountability, especially where safety, compliance, or customer satisfaction is involved.
- Start with a small pilot. Test one workflow with one team before connecting every department.
What to Check Before You Start
Before introducing an AI-assisted workflow, ask yourself four questions. First, do you have enough reliable data to make a useful comparison? Second, are your staff following the same process today, or does each person work differently? Third, what should happen when the system is uncertain? Fourth, who will review and improve the rules over time?
You should also consider data protection. If your workflow includes customer details, employee information, product designs, or supplier documents, control who can access the records. Keep audit trails so you can see what was changed and by whom. A faster process is not useful if it creates confusion about responsibility.
For many SMEs, the first implementation does not need to involve custom AI development. A connected form, workflow automation platform, barcode scanner, image upload, or dashboard may be enough to prove the idea. Once you understand which data is valuable and which exceptions matter, you can decide whether more advanced tools are justified.
The Bigger Picture
Apple’s foldable phone story points towards a broader change in manufacturing and operations. AI will increasingly sit inside ordinary processes, helping workers inspect, compare, predict, and correct. Customers may never see the technology directly, but they will notice its results through more consistent products, fewer mistakes, and faster responses.
That does not mean every Malaysian SME needs a high-tech factory. It means the businesses that document their processes and collect useful operational data will be better prepared to automate them. A company with clear standards can improve step by step. A company that relies only on memory and informal instructions will find automation much harder.
Start by looking for the “hinge” in your own business: the small point where a minor error creates delays, rework, complaints, or wasted staff attention. Measure that step, define the standard, and automate the repetitive comparison. You may not need a dramatic transformation. A carefully improved bottleneck can make your entire operation easier to manage.
Ready to Streamline Your Operations?
Your business should run itself. AutoRunBiz deploys AI agents to automate your daily operations — WhatsApp orders, invoicing, customer follow-ups, and accounting. Book a free 15-min ops audit to see where automation fits your business →