S3.E8 未來的設施 Facility of the future

百靈佳殷格翰正在推動生產系統的數位化，目標不僅是更有效和安全地製造藥物，還能為患者量身定制少量的小包裝。新建造的設施，可以以非常小量生產癌症藥物，代表了這過程中重要的一步。使用虛擬3D模型和擴增實境眼鏡，員工已經在積極規劃這個智慧型工廠。

在工廠裡，一台高速壓藥片機器正在快速製作出很多藥片。這是一個自動測試藥片的機器人手臂嗎？那裡有很多空間可以讓他移動，「我可以很清楚知道在工廠裡移動會是什麼感覺」，Peter Comes博士在拿掉擴增實境的眼鏡時說道。他是百靈佳殷格翰固體生產的負責人，到目前為止，他剛造訪過的超現代工廠只存在於電腦上而已。利用建築師做的模型和機器設計藍圖，數據專家建造了一個動態立體的，目前正在百靈佳殷格翰的總部建造的工廠。憑藉這些VR眼鏡，這個用於生產藥片的新設施進而成真了起來。目前正在Comes辦公室窗口前，建造工程喧囂的進行，這個Launch Facility將從2020年開始為藥品的生產寫下新的篇章。

Launch Facility是百靈佳殷格翰生產數位化的一個典型例子，但它遠超過當初創始的目的。公司總部和其他工廠裡的許多機器早已聯網，中央系統也將一一收集，比較和評估整個生產線的數據。目標是使員工能盡早發現不正常的訊息並解除製造過程中的問題，希望可以在問題發生之前就解決問題，而不是在最終檢查報告時發現藥品嚴重偏離規範內的標準。此時，當IT監督每個製造階段時，最終檢查將變得更加容易。

網路化的生產思維減少了瑕疵數量，使流程更加有效率，就如同質量管理一樣。「我們的行業受到嚴格的監管」，負責公司在德國工廠生產所有人類藥品的AnjaPreißmann博士說道。「我們的生產系統經過德國和外國大量且廣泛審核，效率對我們來說是一個關鍵。」無紙化流程是一個新的例子。以前，主管人員必需手動記錄員工在被允許操作某些類型機器之前已經完成的健康檢查和安全培訓課程。將來，他們將以數位化方式在系統中輸入此類訊息，甚至可以直接將訊息傳送到控制機器的系統，如果操作員不符合其中一個條件，機器甚至不會啟動。

百靈佳殷格翰的生產工程師已經在使用數位化來優化現有設備，以目前的包裝機為例：當新ㄧ批的藥片已準備好進行包裝，機器會發出信號，很多包裝動作隨之進行。但是，專家為其重新設計一條長度超過20米的線路機器，在幾秒鐘內將一個一個藥片裝入之後包裝入袋，最後檢查它們的重量然後裝盒存放和出貨。根據需要被重新調整的機器數量，線路設置可能會花上幾個小時，在此過程中，每個工具的移動都必須精準，這大概需要長達12個月的時間來培訓新員工。

那是機器過去運作的樣子：而今，在其中一生產線上的測試當中，擴增實境的眼鏡幫助新員工完成前置作業。他們通過QR碼自動下載應用程式，每個單獨的步驟都通過眼鏡右上角的棱鏡折射出現在員工的視野中。同時，員工密切關注他的工具和機器，他的雙手可以自由工作。「我們發現這使我們能夠將新員工的培訓時間縮短到六個月左右」，人類醫藥供應德國Preißmann負責人說。

「我們的生產是經過德國和外國大量且廣泛的審核，效率對我們來說是一個關鍵。」

除了這些設置擴增實境的眼鏡包裝機之外，還有第二種以不同模式安裝可用於緊急情況的包裝機。以前，當機器出現問題時，技術人員必須在現場進行檢查，即使在夜班期間也要進入工廠裡。現在，遇到這種情況，員工可以打電話給技術人員並戴上AR眼鏡，通過眼鏡中安裝的鏡頭，技術人員可以在家裡的電腦上看到現場員工所看到的情況，並提供需要的協助。「如此一來，如果我們德國境外的設施出了問題，我們也能夠快速提供援助，而不需要技術人員飛到現場」，Preißmann說。

同樣的技術也適用於雙重查核：員工目前需要親自到現場，以便根據雙重驗證方法進行重要檢查。由於嚴格的安全和衛生規定，這是一個費力的過程，從通過設施內的大門就需要一段時間，而且很多時候都只允許少數人進入特別敏感的區域。使用AR眼鏡，員工將很快就能夠從辦公室進行這些檢查。該公司的生產專家目前甚至考慮在收貨區域安裝創新的區塊鏈技術，相機可以掃描每個包裹並與區塊鏈數據庫中的代碼作相對應檢查。這將使購買小批量包裝材料更有效。「針對我們生產系統上的數位化轉型，我們已經啟動了許多計劃」，Preißmann說。

「一旦為患者客制化藥物不再是例外，我們就會做好準備將它變成常態」
──Anja Preißmann博士

此外，Preißmann和她的同事們正在開啟一個全新的世界：未來的包裝機器將不再像現在這樣單獨的在固定線路上運作，而是將以網絡的方式運行。填裝，印刷和包裝站將透過數位化技術，按照各種藥品、數量和不同的市場來生產包裝，如果需要，將會添加更多的機台。人類專家和合作的機器人在各個階段上一同工作，且自動化水平可以自由選擇。

「透過新設施，我們的目標是以大批量的生產成本來生產單獨的包裝尺寸」，Preißmann說。機械工程師目前正與該公司的專家密切合作，開發和建造這個設施及其零件。「這種組合式的和自動化生產是製藥行業中的一項創新」，Preißmann說。

事實上，員工今天已經可以使用VR眼鏡查看Launch Facility，這不僅是一種噱頭，目的是幫助在工廠完成之前改進流程：任何查看工廠電腦模型的人都能夠提出改進的建議。「這為我們提供了寶貴的建議」，Peter Comes博士說。在規劃階段落實建議比在建築完工後更容易也更便宜。 Comes表示，這也將促進機器操作員對新工作場所的熟悉度：「畢竟生產線將根據他們的想法進行建構。」

回到包裝上的話題，百靈佳殷格翰在臨床測試中需要比較少量批裝小的藥品，例如，當新開發的藥物需要經歷各種測試階段的時候。測試時，開發人員通常只需要幾片新藥，隨著測試數量增加與時間推移，需要的量將會更多。然而，小批量也越來越成為常規生產的特徵，即使是現在，百靈佳殷格翰在腫瘤學領域的產量也非常小。癌症藥物通常不是大規模生產的藥物，並且部分僅根據特定國家的特定患者之訂單進行包裝。因此，數量是微不足道的，只是剛好所需的數量而已。

「在規劃階段落實建議比在建築完工後更容易也更便宜。」
──Peter Comes博士

Boehringer Ingelheim is pushing forward with the digitalisation of its production systems. The goal is to manufacture medicines not only more efficiently and safely, but also in ever-smaller batches tailored to individual patients. The construction of the new Launch Facility – which can produce cancer medicines, for instance, in very small batch sizes – represents a major step on this path. Using a virtual 3D model and virtual reality goggles, employees are already helping to plan for this smart factory.

H ere a tablet press emitting tablets at high speed. Is that a robot arm over there automatically testing single tablets? There is a lot of space to move around in between. “I have a very good sense of what it will be like to move around the plant,” says Dr Peter Comes as he removes his virtual reality goggles. He is Head of the solids production in Ingelheim. So far, the ultramodern manufacturing facility which he has just visited only exists on a computer. Using the architect’s model and design drawings for the machines, data specialists have created a dynamic, three- dimensional copy of the Launch Facility which is currently being built at Boehringer Ingelheim’s headquarters. But with these VR goggles, this new facility for the production of tablets comes alive even now. Where construction machinery currently is making a din in front of Comes’ office window, the Launch Facility will mark a new chapter in pharmaceutical production from 2020 onwards.

The Launch Facility is a prime example of the digitalisation of production at Boehringer Ingelheim, but it is far from the only initiative of its kind: many

machines at the company’s headquarters and other facilities have long since been networked, and a central registration system will soon collect, compare and evaluate data for entire production lines. The goal is to enable employees to identify irregularities and to eliminate problems in the manufacturing process as early as possible – possibly before they occur and not only when the final inspection reports that medicines deviate more strongly from the norm than the strict tolerance limits permit. At the same time, this final inspection will become significantly easier in future when IT supervises each individual manufacturing phase.

Networked production thinking reduces the volume of rejects and makes processes more efficient, just like quality management. “Our industry is strongly regulated,”says Dr Anja Preißmann, who is responsible for the production of all human pharmaceuticals at the company’s German facilities. “Our manufacturing system undergoes a large number of extensive audits conducted by German and foreign authorities. Efficiency is a key issue for us here.” One example of the new, paper-free processes is that shift supervisors were previously required to document manually which health checks and safety training courses an employee had completed before being permitted to operate certain types of machinery. In future, they will enter this type of information in the system digitally and can even directly link it to the machine’s control system: if an operator does not fulfil certain preconditions, the machine will not even start.

Boehringer Ingelheim’s production engineers are already using digitalisation in order to optimise existing equipment. To take the example of packaging machines: there is currently a lot of movement once the signal is given that a new batch of tablets is ready for packaging. Specialists then refit one of the lines, which is more than 20 metres in length, for the new job. The machines subsequently fill the blisters with tablets in a matter of seconds, push them into packages together with the package insert, and    finally check their weight before they land in boxes for storage and dispatch. Depending on how many of the individual machines need to be reorganised, the set-up of a line can take several hours. During this process, every movement with each tool has to be exact. It takes up to twelve months to train new employees.

At least that’s the way it used to be: in a pilot project on one of the lines, augmented reality goggles are helping new members of staff with the set-up process. They automatically load the appropriate set-up programme by means of a QR code. Each individual work step subsequently appears in the employee’s field of vision via a prism in the top right-hand corner of the goggles. At the same time the employee keeps an eye on his tools and machines and his hands are free for work. “We have found that this enables us to reduce the training period for new employees to six months,” says Head of Human Pharma Supply Germany Preißmann.

“Our manufacturing system undergoes a large number of extensive audits conducted by German and foreign authorities. Efficiency is a key issue for us here.”
DR ANJA PREISSMANN

In addition to these augmented reality goggles for the set-up of packaging machines, a second, differently configured model is available for emergencies. Previously, when a machine experienced a problem, a technician had to carry out an inspection on site, coming into the facility even during a night shift. Now, in that type of situation the employees can call the technician and put on the AR goggles. Through a camera fitted in the goggles, the technician can see on his laptop at home what the employee on site can see and provide instructions on what needs to be done. “This way, in future we will also be able to provide rapid assistance if our facilities outside Germany run into problems, without the need for a technician to fly over there,” says Preißmann.

The same applies for double checks, where employees currently visit in person on site for important inspections in line with the double-verification approach. This is a laborious process due to strict safety and hygiene regulations: it takes a while to get through the gates inside the facilities, and only a few persons are permitted to enter particularly sensitive areas at any one time. With AR goggles, one of these two employees will soon be able to carry out these checks from the office. The company’s manufacturing experts are currently even considering installing innovative blockchain technology in the goods receipt area. Cameras could scan package deliveries and check their codes against a blockchain database. That would make the purchasing of small batch sizes of packaging material more efficient. “For the digitalisation of our production system, we have already started a number of initiatives,” says Preißmann.

In addition, Preißmann and her colleagues press on into an entirely new world: packaging machines will in future no longer operate along a fixed line as they currently do and will instead work as individual, networked modules. The filling, printing and packaging stations will be digitally configured for various medicines, quantities and different markets. Further stations will be added, if required.  Human specialists and collaborating robots work together on the individual modules, and the level of automation can be freely selected.

“With the new facility, our goal is to produce individual packaging sizes at the same cost as huge batches,” says Preißmann. A mechanical engineer is currently developing and building this modularised facility and its components in close cooperation with the company’s experts – this is an individual item. “This type of modularised and automated manufacturing is a real innovation in the pharmaceutical industry,” says Preißmann.“

The fact that employees can already view the Launch Facility today using VR goggles is not just a gimmick, by the way. It is intended to help to improve the processes even before the plant has been completed: anyone who has looked around the computer model of the facility will be able to put forward suggestions for improvements. “That provides valuable feedback for us,” says Dr Peter Comes. It is much easier and cheaper to implement proposals during the planning stage than later on in the finished building. This will also boost the machine operators’ attachment to their new workplace, according to Comes: “After all, it will have been built in line with their ideas.”

Back to packaging: Boehringer Ingelheim requires small batch sizes during clinical tests, for instance, where newly developed medicines undergo various testing phases. Developers typically require just a few tablets of a new medicine and then ever more of them over time, as the number of test subjects increases. However, small batches are also increasingly an everyday feature of regular production: even now, Boehringer Ingelheim produces very small quantities in the area of oncology. Cancer medicines are not mass-produced goods and are in part only packaged on the basis of an order for a specific patient in a particular country. Accordingly, the number of units is minuscule – it’s precisely the quantity required.

This type of order is currently handled via fully automated mass production packaging lines, since individual work steps such as packing tablets in blisters cannot be outsourced. This means that technicians spend several hours refitting the machines. They then run the equipment for this very small order, even if this is just for a few minutes. “In future, we will package this type of order in the new modular packaging line,” says Preißmann. “We will thus be ready once individualised medicines for patients are no longer an exception and become the norm.”

“It is much easier and cheaper to implement proposals during the planning stage than later on in the finished building.”
DR PETER COMES