It was Monday evening at the office. All the personnel, except Juha Kalliopuska, co-founder/CEO of Advacam Oy, had already left home. I was writing a project application, and the moment was perfect for it, because it was the most peaceful time of the day. I could only hear the smooth humming of my laptop, but there was no disturbing noises. The silence was occasionally broken by the sound e-mail alerts. I was in a good flow for writing, so I didn’t want to distract my mood by checking them. After a while, Juha appeared at the door and said “Good news from CERN! Have you checked your e-mail?”“. It turned out that Advacam was granted the whole pixel detector hybridization work for the CERN LHCb VELO experiment. It had already been a good day, so this news was the cherry on top of the cake. The deal was not only good for us in terms of production volume, but it was also the first call for tender related to experiment upgrades, which we won independently. We have a long collaboration with CERN, but it still took time to earn CERN’s trust and become an exclusive supplier for goods within a major high-energy physics experiment.
Let’s take a few steps back and see how Advacam’s relationship with CERN has developed over the years. It was March 2003 when I started to work for VTT, the technical research centre of Finland. I graduated from Tampere University of Technology in 2002, and I was lucky to find a job that corresponded to my education. I was happy to begin working on the flip chip process and generic thin film technologies in a clean room environment, because there weren’t many companies in that field in Finland. When I joined VTT, they had recently begun collaborating with CERN, providing flip chip solutions for assembly of pixel detectors for high-energy research purposes. The pixel detector concepts had been demonstrated at CERN in the 90’s, and CERN wanted to implement the noise-free detector technology for the inner layers of the vertex detectors in Large Hadron Collider (LHC) experiments. After years of collaboration, VTT successfully delivered the pixel detector modules for ALICE and LHCb rich experiments. The productions done for CERN were important milestones for VTT, but also for the beginning of Advacam, because personal relationships were made.
At VTT, we had both the flip chip process and silicon (Si) sensor fabrication process, which were a very promising technology combination. Still, it took many years for the collaboration to really flourish. Because of this collaboration with CERN, I was able to participate in the CERN fellowship programme for two years. During the programme, I learned to understand the needs of CERN very well, and I met many technical people working with the CERN experiments. This time at CERN (2008 – 2010) was very valuable for me, because I could also deepen my knowledge on the technology, as well as the groups working with pixel detectors. While I was at CERN, Juha had made good progress with the Si sensors processing at VTT, and he had successfully implemented the fabrication process flow for the so-called edgeless Si sensors (i.e., sensor without guard rings). The edgeless sensor technology was the first connecting line between the current Finnish and Czech shareholders of Advacam.
“There were not many connecting aspects in the beginning; all we knew was that both the Finns and Czechs like beer and ice hockey…”
When returning from CERN to VTT, I started as a team manager for the 3D integration team. At that time, the synergies were found in between the Si sensor and the micro packaging team. The combination of the two technologies was very unique and even today there are not many companies in the world offering a similar service for hybrid pixel detectors. Despite of the valuable technology at VTT, the implementation of the combined technologies wasn’t successful, because the team wasn’t simply strong enough. This was also the reason why Advacam was founded in 2012 as a technology spin-off company of VTT. We saw the high potential for these technologies, but the constellation at VTT was not right. By getting the right people together, we could make products from detector designs to complete X-ray camera products. Advacam s.r.o (former Widepix s.r.o.) was founded one year later joining the Finnish interest company Advacam Oy. It was difficult to find connecting aspects in the beginning; all we really knew was that both Finns and Czechs like beer and ice hockey. Still, it was important to get the Czech people on board, because they had been actively developing X-ray imaging applications (electronics, software, algorithms, system and many more) around the pixel detector modules, and they were widely known for their talent. Today, both sides of the company are under the board of directors and they have also been boot strapping, successfully avoiding having to raise venture capital funding.
Since the beginning of Advacam, we have been working with dozens of projects related to the research done at CERN. We have the capability to produce exotic Si sensor and we have highly specialised talent in the flip chip bonding of pixel detectors. Although we were a tiny company, we offered interesting services for CERN, and because of the long personal history, we were able to gain their trust. The requirements set by CERN are rigorous, because of the need to have thin Si pixel detector modules, which eliminates a great number of disturbing collisions and noise in the measurements caused by the excess Si material. One of the major milestones for us was the CMS upgrade project done together with Helsinki Institute of Physics. This was Advacam’s first real production project for CERN and we received the big order with the help of HIP. Despite the occasional technical difficulties, we were successful with the work and we provided the best detectors in terms of fabrication yield. Now, several years later, we left a bid in call for tender for flip chip processing of pixel modules for the VErtex LOcator (VELO) part of the LHCb experiment at CERN. We expected to enter the qualifying part of the project, which was a quality-based milestone for the actual production project. However, it turned out we received the whole deal, which felt special, because it meant that we had established an even higher-level of trust with CERN.
Why is trust important? The investments in the experiments at CERN are huge. The experiments are very complicated systems, which are difficult to perceive even for physicists. The LCH experiments are using the same beam line and there are set time windows for each experiment for performing it measurements or to have the upgrades installed. When the beam line is off during the construction of the apartment building sized experiments, no particles will be collided. In this respect, it is very important for the suppliers to be able to deliver all the necessary goods well in advance. If the deliveries cannot be made within reasonable time, it could compromise the construction and inauguration of the whole experiment, which has been designed to last for five to ten years. Another layer of the complexity is the requirement of highly reliable components to be mounted in the detector structures. The experiment will be run for years, before there is a chance to dismantle and change the detector structure. In this respect, the reliability requirements are comparable to space technologies.
The detectors required by CERN are not bulky, but very challenging structures to be produced in rather high volumes for a rather short time. For the industrial partner, it’s also not easy, because they’ll have to set the manufacturing processes for products that have no long-term future. These kinds of intensive projects often distract from other company activities, but the timing was perfect for us, and we had the capacity on our fabrication line. At the same time, CERN wants to invest in technology that will remain on the market for its need in the future. In this respect, it is important for us to have pixel detector product technology, which will also support CERN in the future.
In larger upgrade projects, such as ATLAS and CMS, the qualification process of the suppliers is much tougher than in case of the LHCb VELO. One has to deliver similar prototype modules of the detectors to be used in the final experiment. The modules will be characterised at CERN and it will affect the selection of suppliers. The manufacturing of a few prototypes is very expensive when Si sensor fabrication is included. The construction of the prototype modules will be funded by the potential supplier. It is a major and risky contribution, when there no certainty of receiving an order from CERN. Moreover, the supplier must have had certain sales of similar Si detectors for a defined value in order to qualify the bidding.
This story shows how a small start-up company can do business with CERN. When working with CERN, the common history plays a major role. However, I consider that it is even more important to have an innovative team that pushes the technology barriers forward, because of the high technological discipline level of CERN.
Sami Vähänen, CTO of Advacam
