Üretimde hız baskısı artıyor, toleranslar daralıyor, enerji maliyeti yükseliyor; aynı hatta hem kaliteyi korumak hem de güvenliği artırmak artık “fazla mesai”yle çözülmüyor. Bu noktada R&D ve inovasyon, endüstriyel otomasyonun sadece daha hızlı değil, daha öngörülebilir ve ölçülebilir çalışmasını sağlıyor.

By 2026, AI-powered predictive maintenance, IoT-based real-time monitoring, robotic applications and sustainability-focused improvements will come to the fore in Turkey. To give a simple example, when mould temperature and cycle time are kept stable using sensor data, scrap rates fall, downtime decreases and operators need to enter hazardous areas less often.

In this article, we will examine how R&D is implemented in the field, at what stage innovation generates real value, and its impact on production speed, quality, cost, and workplace safety, using clear examples. You will also explore Erdem Makina's approach not through "brand rhetoric" but through problem definition, trial and error, data-driven decision-making, and standardisation logic. For a related framework, the content of Industry 4.0 and aluminium casting innovations also provides a good starting point.

The tangible contribution of R&D in industrial automation: speed, quality and less downtime

R&D is the bridge that translates the desire for "faster operation" in automation into measurable results. What we refer to as speed in production is often the cycle time, i.e., how many seconds it takes for a part to be produced from start to finish. Quality, on the other hand, is less about "looking good to the eye" and more about tolerances verified by measurement and a low scrap rate. Downtime is the period during which the machine stops production, meaning lost parts and lost energy every minute. When R&D is done correctly, these three areas improve simultaneously because the process becomes more stable.

Why does automation quickly become stagnant without R&D?

Everything looks fine during the initial setup, but problems accumulate as the process grows. The most common bottleneck is rising maintenance costs. This is because the system is patched with a "just keep it running" mentality rather than addressing the root cause of the failure. This translates to more frequent downtime, more spare parts, and more service calls.

The second issue is old control systems. If the PLC, drives or operator panel are not up to date, software changes become difficult. Even the smallest revision is postponed for fear of "the line stopping". Over time, the line becomes unable to adapt to new products or new quality standards.

Third heading spare part availability. A module that is not available on the market can turn a minor fault into days of downtime. R&D here means not only part selection, but also alternatives for critical parts, standardisation and documentation.

Finally, operator errors arise. Complex screens, ambiguous alarms and non-standard settings cause the same task to be performed differently in each shift.

The critical difference here is this: Copying the current system also copies the same errors. Improving the process adds measurement, strengthens control, reduces human dependency and prepares the system for the future.

The practical outcome of innovation: fewer errors, more stable production

In most facilities, innovation is understood to mean "buying new machinery". Yet its real-world equivalent is simpler: the right sensor, the right alarm, the right control.

For example, sensor selection. If you use the wrong sensor in the wrong place or of the wrong type, the system becomes blind. The right sensor, however, regularly collects data such as temperature, pressure, flow, and vibration; the process repeats the "same condition". This means stable production.

When translated into everyday language, automatic control means this: The operator should not have to make adjustments manually each time; the system should maintain the target value itself. Data tracking, on the other hand, is "looking back at past data to find the cause when a problem occurs."

A clear example of starting small and reaping big benefits: Adding an additional measurement point at a critical point and defining an automatic alarm linked to it. Let's say the lubrication pressure is dropping. If the operator notices this too late, part quality will deteriorate and the machine may stop. If the alarm is triggered early, the downtime is short, and intervention occurs before scrap is produced. This approach also reduces errors in robotic applications; for example, in systems such as Precision production with linear casting robots, repeatability, accurate monitoring, and control achieve their true value.

How do you measure the success of R&D? 6 indicators to track

It is not enough for R&D to "feel good"; the numbers must speak for themselves. The following six indicators clearly demonstrate the impact on speed, quality and posture:

1. Availability (Part 1 of OEE): Did the machine actually run for the planned time? It increases if downtime decreases.
2. Performance (the second part of OEE): Did the machine produce at target speed? It increases as cycle time improves.
3. Quality (the third component of OEE): How many of the produced parts came out flawless? It increases as scrap and rework decrease.
4. Failure frequency: How often do failures occur? Frequent failures indicate poor design or incorrect maintenance.
5. MTTR (mean time to repair): How quickly do you recover when a fault occurs? Well-designed automation enables rapid fault diagnosis.
6. Energy/product: How many kWh do you consume to produce a part? Monitoring and process stability reduce unnecessary consumption.

By regularly monitoring these indicators, the contribution of R&D becomes clear: shorter downtime, lower scrap rates, and more predictable production.

Erdem Makina's approach: transforming problems arising in the field into solutions through engineering

Problems encountered on the shop floor usually manifest themselves through noise; downtime increases, quality fluctuates, and the operator cannot maintain the "same setting". The solution drawn up at the drawing board sometimes fails to deliver in real life. This is where Erdem Makina's approach differs: it identifies the problem not when the fault occurs, but within the production flow, measures it, clarifies the root cause, and then transforms it into a permanent solution through engineering. This is not a one-off "improvement project"; it means establishing a repeatable R&D routine.

How is an R&D culture established? Team, testing framework and clear objectives

An R&D culture is born not from grand statements, but from a small yet disciplined framework. The first step is for everyone to speak the same language. Maintenance says "breakdown", production says "downtime", quality says "scrap"; R&D reduces these three words to a single question: Where, when, and under what conditions does the problem occur?

A practical structure that works for this is as follows:

1. Problem pool: Field teams record "burning" issues. A short format is sufficient: date, line, symptom, impact (downtime, number of scrap items, occupational safety risk).
2. Root cause analysis: Symptoms alone are not sufficient. Saying "the sensor is malfunctioning" is a result; the real question is "why is it malfunctioning?" Issues such as incorrect installation location, temperature effects, cable routing, software filters, vibration, and operator habits are examined.
3. Prototype and controlled trial: The solution is first tested on a small scale. The aim is to validate the idea without putting production at risk.
4. Field test: The test is conducted in real-world conditions. This is because certain faults only manifest themselves in the presence of heat, dust, shift patterns and the actual flow of parts.
5. Revision and standardisation: This is the most critical point. If the solution works, it should not remain as the "master's secret setting"; it must become standardised.

There are two things that make this culture sustainable: documentation and standardisation. Documentation eliminates the "who remembers" problem during the next breakdown. Standardisation ensures that the same language is spoken on the same type of machine; details such as alarm names, cable labels, spare parts lists, and software versions directly reduce MTTR. In short, R&D favours good repetition, not good ideas.

Expand the project with university and institutional support (e.g. TÜBİTAK)

Some problems can be solved using internal company resources. Others require longer testing, more measurements, and more rigorous reporting. Support from organisations such as TÜBİTAK-TEYDEB comes into play precisely at this point. The greatest contribution of such programmes is not so much the "money" as the discipline; it clarifies the objective, the method, and the output.

The main areas covered by the support provided are as follows:

• Budget and resource plan: This creates a more comfortable framework for items such as sensors, test equipment, software development, and validation work.
• Testing and verification approach: Merely stating "it works" is insufficient; progress is made using a measurement plan, acceptance criteria, and comparison method.
• Reporting discipline: Regular reporting creates a shared memory within the team. It also speeds up subsequent projects.

So what types of problems are more suitable for a supported project? A supported model makes sense if three criteria are met: (1) if the impact of the problem is high (downtime, energy, quality), (2) if the solution involves multiple disciplines (mechanical, automation, process, software), (3) if long-term field data is required for verification. For example, control strategies that increase process stability, closed-system designs that reduce energy consumption, and safety and automation improvements that reduce human intervention fall within this scope.

The fact that Erdem Makina's completed projects include work deemed worthy of support by TÜBİTAK-TEYDEB is clearly stated on the official website, demonstrating that this approach has been put into practice. To follow this type of project and R&D content, the R&D and projects page on the Erdem Makina blog provides a good reference area.

The focus visible on the R&D page: continuous investment and innovation logic

Sustainable R&D is only possible with a consistent investment approach, not with a "let's do something this year" motivation. The key point in Erdem Makina's R&D narrative is its approach of investing in R&D and innovation every year. This ties into a continuous improvement cycle rather than postponing problems in the field to the following year's budget.

It is also clear that R&D is not solely a closed-door affair: alongside engineers and technical staff, university collaborations are emphasised. The university side adds depth, particularly in areas such as measurement methods, experimental design, and material and process knowledge. The field side, on the other hand, takes the idea of "working in the laboratory" and makes it workable in production. When these two do not come together, the result is either expensive but fragile solutions or cheap but temporary fixes.

The practical application of innovation here is as follows: designing, testing, validating and standardising in the field according to customer needs. If you wish to see the main framework of the R&D approach directly from the source, Erdem Makina’nın Ar-Ge yaklaşımı sayfası bu süreklilik, üniversite iş birlikleri ve destekli projeler vurgusunu bir arada sunar.

From concept to implementation: the innovation process in automation projects, step by step

Bir otomasyon projesinde inovasyon, “yeni bir şey yapmak” kadar, doğru sırayla ilerlemek demektir. Sahada her dakika üretim akıyor, duruşun ve hatanın maliyeti büyüyor. Bu yüzden fikir aşamasından devreye almaya giderken yalın bir yaşam döngüsü kurmak gerekir: ihtiyacı netleştir, küçük ölçekte dene, sahada güvenli şekilde devreye al, stabilize et. 2026’da Türkiye’de öne çıkan yaklaşım da buna yakın, robotik, sensörler ve veri takibinin birlikte düşünülmesi, kararların veriyle desteklenmesi, enerji ve kalite hedeflerinin aynı tabloda ele alınması bekleniyor.

Bu akış oturunca “en pahalı çözüm” tuzağından çıkarsın. Çünkü pahalı olan çoğu zaman ekipman değil, yanlış problem için yapılan yatırımdır.

İhtiyaç analizi: doğru problemi seçmek, yanlış yatırımı önler

İyi bir ihtiyaç analizi, otomasyonun en kritik sigortasıdır. İlk adım “ne yapacağız” değil, hangi problemi çözeceğiz sorusudur. Örneğin hedefin çevrim süresini kısaltmak mı, hurdayı azaltmak mı, operatörü riskli alandan çekmek mi, yoksa enerji tüketimini kontrol altına almak mı? Bu hedeflerden biri netleşmeden yapılan proje, sahada sürekli revizyona döner.

Burada pratik bir çerçeve işe yarar:

• Problem tanımı: Belirtiyi değil kök etkiyi yaz. “Robot sık duruyor” yerine “X istasyonunda kavrama hatası nedeniyle vardiya başına Y kez duruş” gibi.
• Kapsam: Hangi makine, hangi istasyon, hangi ürün grubu? Kapsam büyüdükçe risk ve süre uzar.
• Başarı ölçütleri (KPI): Projeyi bitirince neyi ölçüp “tamam” diyeceksin? OEE, çevrim süresi, hurda oranı, MTTR, enerji/ürün gibi metrikleri baştan seç.
• Bütçe ve süre beklentisi: “Ne kadar sürede geri dönecek” sorusu burada cevaplanır. Bütçeyi kalem kalem yazmak, sürprizi azaltır.

En çok fayda sağlayan çözüm mantığı şudur: Yatırımın büyüklüğünden önce etki alanına bak. Bir sensör ekleyip doğru alarm kurgulamak, bazen yeni robot almaktan daha çok duruş düşürür. Benzer şekilde, prosesin stabil olmadığı bir hatta “daha hızlı” otomasyon kurmak, sadece hatayı hızlandırır. Bu yüzden analiz aşamasında kısa bir “mevcut durum fotoğrafı” çekmek şarttır: son 3-6 ay duruş kayıtları, hurda sebepleri, vardiya farklılıkları, kritik güvenlik riskleri.

Süreç ve kalite dayanımı gibi temel hedefler, döküm hatlarında da belirleyicidir. Bu perspektifi genişletmek istersen, Alüminyum dökümde dayanıklılık artırma yaklaşımı, proses istikrarının otomasyondaki etkisini iyi anlatır.

Tasarım ve prototip: küçük dene, hızlı öğren

İhtiyaç netleşince tasarım başlar, ama tasarımın amacı “en şık çözüm” değildir. Amaç, en düşük riskle hedef KPI’lara yürümektir. Bu yüzden prototip, otomasyon projelerinde bir masraf değil, maliyeti düşüren bir erken uyarı sistemidir.

Prototipi basit düşün: Üretimi durdurmadan, gerçek hat koşullarını taklit eden küçük bir deneme.

• Simülasyon: Robot erişimi, çevrim zamanı, çakışma riskleri gibi konuları bilgisayar üzerinde görmek.
• Küçük test standı: Bir kavrayıcı, bir sensör, bir aktüatör ve temel yazılım. Amaç “çalışıyor mu”yu görmek, detayları değil.
• Pilot hat: Tüm sistemi ana hatta taşımadan önce dar kapsamlı bir istasyonda denemek.

Bu aşamada en çok kazandıran şey, “sürprizleri” erkenden yakalamaktır. Örneğin parça toleransı sahada simülasyondan farklı çıkabilir, ortam sıcaklığı sensör okumalarını kaydırabilir, operatör alışkanlığı iş akışını değiştirebilir. Prototip bunları üretime yaymadan gösterir.

Prototipte hedefi tek cümleye indir: Riski erken gör, revizyonu ucuzken yap. Sonra tasarımı buna göre olgunlaştır. 2026 trendlerinde de bu bakış güçleniyor; sensör verisini ve izlenebilirliği baştan kurgulayan tasarımlar, sonradan eklenen çözümlere göre daha hızlı toparlıyor.

Kalıp yüzey işlemleri gibi tekrarlanabilirlik isteyen işlerde, küçük denemeler özellikle fayda sağlar. Örneğin otomatik spreyleme gibi alt süreçlerde, ayarların standartlaşması hem kaliteyi hem de tüketimi etkiler. Bu başlıkta Cost savings in pneumatic mould spraying Focused examples embody the principle of "small change, big impact".

Field tests and commissioning: safety, training and stabilisation

Commissioning is the most visible moment of the project, but it also carries the most risk. That is why good teams manage commissioning not as a matter of "let's go and connect it so it works", but as a planned transition. Security, training and the first two-week stabilisation period in particular determine the project's longevity.

A successful commissioning plan includes the following components:

1. Commissioning plan and downtime window: Which day, which shift, in what order? Is a rollback scenario defined?
2. Safety check: Emergency stop, lockout/tagout (LOTO) procedures, safety relays, safety sensors, access to hazardous areas. Safety is not a "final check"; it is part of the test plan.
3. Operator training: Screens, alarms, product changeover, basic troubleshooting steps. If the operator understands "why it stopped", they will respond correctly.
4. Common language with the maintenance team: Alarm names, I/O list, spare part codes, software version. This alignment directly reduces MTTR.
5. Spare parts list: Critical sensors, valves, drivers, connection elements. If "availability" is lacking, even the best system will stop working.
6. Acceptance tests (FAT/SAT logic): Target KPIs, cycle time, security scenarios, quality measurement, traceability. If the acceptance criteria are written down, there will be less debate.

The first two weeks after commissioning are truly a stabilisation period. Minor adjustments are normal during this period and should even be expected. The important thing to note here is to maintain the distinction between "making adjustments" and "shifting scope". A short daily routine is useful for managing stabilisation: the previous day's downtime, the 3 most frequent alarms, the initial response time, scrap causes and operator feedback. With this data, software filters, sensor thresholds, robot speed profiles and process parameters are collected in a controlled manner.

Ultimately, the goal is clear: it is implemented not just for the system to work, but to work stably. Once this discipline is established, innovation in the field becomes not a one-off project, but a recurring cycle of improvement.

January 2026 trends: 5 topics shaping the R&D agenda in automation

As we enter January 2026, the main drivers of R&D in automation remain unchanged: cost pressure, quality fluctuations and downtime risk. What has changed is that the approach to these issues has become more measurable. Trends are not centred around "more data" or "newer devices", but rather around choosing the right technique for the right target.

The five most popular topics in the field are: energy efficiency, traceability, cyber security, flexible production, and the edge computing approach, which reduces latency. Edge computing processes data at the machine itself, without always transferring it to the cloud, thereby reducing response times and network load. The following four topics are the most frequently "translated into business" part of the R&D agenda.

Energy efficiency and sustainable production targets

As energy costs rise, the direction of R&D becomes clear: achieving the same production with fewer kWh. The most practical gain here is not to monitor energy consumption with a single meter, but to view it on a station-by-station basis. This is because losses often occur not while the machine is running, but during idle periods, unnecessary heating, or incorrect pressure settings.

There are three levers frequently used by R&D in the field. The first is energy monitoring: by tracking major consumers such as presses, furnaces, compressors, and cooling systems separately, it becomes clear which step is inflating costs. The second is motor drives (VFD/servo): speeding up or slowing down a fan or pump rotating at a constant speed according to need makes a significant difference, especially at partial load. Thirdly, process optimisation: shortening the cycle time alone is not sufficient; it is necessary to narrow and stabilise the temperature, pressure, and time windows.

This approach also has a direct impact on robotic applications. For example, correct synchronisation and the right parameter set both shorten the cycle and reduce unnecessary air and energy consumption. To see how robotic solutions are linked to process discipline, the system logic on the page Industrial automation with aluminium injection robots provides a good example.

Data collection and traceability: seeing without failure

Traceability is not about receiving reports after a fault has occurred. The aim is to catch the warning signs that give a "low-level alert" before a fault occurs. The way to do this is not to "collect data", but to collect the right data at the right frequency. Recording every signal in milliseconds only generates noise in most facilities.

A simple setup will suffice: sensor data (temperature, pressure, current, vibration), associated alarm logic and trend screens that the operator can understand. Dividing the alarm into two parts yields practical results: "stop immediately" alarms and "monitor" warnings. For example, if the lubrication pressure drops suddenly, you first give a warning and show the trend; if the drop continues, the shutdown kicks in. This way, both unnecessary downtime is reduced and the real risk is not missed.

The gain in terms of R&D is clear: everyone makes the same comment on the same error, the root cause is found more quickly, and MTTR decreases.

Cyber security and reliability: basic measures to prevent downtime

Cybersecurity in production is often not as simple as "preventing data theft"; the real risk is the problem of downtime. Therefore, without getting bogged down in technical details, a practical set of fundamentals is useful.

A brief practical checklist:

1. Network segregation: Separate the production network from the office network, provide remote access from a controlled point.
2. Authorisation: Define role-based users in PLC, HMI and SCADA, remove default passwords.
3. Backup: Regularly back up the PLC programme, HMI project and recipes, and perform a restore test.
4. Update plan: To manage the fear that "it will stop when updated", implement a planned downtime window and testing step.

These four steps significantly increase reliability in most facilities.

Flexible production: lines that quickly adapt to different products

Flexibility is not just a concern for large factories. In small businesses, the product mix changes more rapidly; one day it's a short run, the next day a different part arrives. Therefore, the goal of R&D should not be to completely overhaul the line, but to make the change quickly and safely.

Three practical solutions stand out here. Modular fixtures enable different parts to be attached at the same station with minimal mechanical intervention. Quick recipe changeover reduces errors (incorrect settings, wrong product) by allowing the operator to select the correct parameter set from the screen. Standard parts simplify maintenance and stock management; the same sensor, the same valve, the same driver family means shorter downtime.

The impact of flexibility on the business is measurable: product changeover times are reduced, first-part approval is accelerated, and operator dependency is reduced. R&D here establishes a more repeatable system, not a more complex one.

Conclusion

In industrial automation, R&D and innovation guarantee stable production as well as rapid production. On the 2026 agenda, robotic integration, AI-driven analysis, traceability and energy efficiency are all working towards the same goal: fewer stoppages, lower scrap rates and clearer decisions.

The Erdem Makina approach is a method that converts problems arising in the field into measurable targets, verifies them through small trials, refines them with field feedback, and standardises them. This discipline removes the project from equipment procurement and converts it into a repeatable improvement cycle.

Key findings:

• Set measurable targets, select KPIs from the outset (OEE, scrap, MTTR, energy/product).
• Start small, spot the risk early.
• Integrate field feedback into the daily routine (standby, alarm, scrap reason).
• Include operator and maintenance training as part of the commissioning plan.
• Maintain standards with documentation, spare parts and version management.
• Lock continuous improvement into the monthly review meeting.

What can you do tomorrow?

• Identify the three most common reasons for standing in the last 30 days.
• Select a single sensor or alarm enhancement for each one.
• Set a two-week pilot target and report the result numerically.

If you are seeking rapid gains on the moulding side, you can examine examples of cost savings with pneumatic mould spraying technology and adapt them to your own line.