Sunday, Apr 14, 2024
Update At 14:00    USD/EUR 0,92  ↑+0.0002        USD/JPY 151,69  ↑+0.174        USD/KRW 1.347,35  ↑+6.1        EUR/JPY 164,16  ↑+0.143        Crude Oil 85,49  ↓-0.76        Asia Dow 3.838,83  ↑+1.8        TSE 1.833,50  ↑+4.5        Japan: Nikkei 225 40.846,59  ↑+448.56        S. Korea: KOSPI 2.756,23  ↓-0.86        China: Shanghai Composite 3.015,74  ↓-15.745        Hong Kong: Hang Seng 16.512,92  ↓-105.4        Singapore: Straits Times 3,27  ↑+0.018        DJIA 22,58  ↓-0.23        Nasdaq Composite 16.315,70  ↓-68.769        S&P 500 5.203,58  ↓-14.61        Russell 2000 2.070,16  ↓-4.0003        Stoxx Euro 50 5.064,18  ↑+19.99        Stoxx Europe 600 511,09  ↑+1.23        Germany: DAX 18.384,35  ↑+123.04        UK: FTSE 100 7.930,96  ↑+13.39        Spain: IBEX 35 10.991,50  ↑+39.3        France: CAC 40 8.184,75  ↑+33.15        

i-PRO: Modularization key to developing innovative products as AI transforms video surveillance

Interview - December 19, 2023

In 2022 Panasonic i-PRO Sensing Solutions changed its name to i-PRO when it became independent from its parent company. More than a brand change, this move represented a turning point that redefined the company, reconnected relationships, and reformed its organization – all with an eye on the shift driven by artificial intelligence (AI).


Over the past 25-30 years, Japan has seen the rise of regional manufacturing competitors who have challenged Japan’s dominance when it comes to consumer electronics, but Japan is still the leader when it comes to B2B industrial applications. How have Japanese firms maintained this leadership and what are the unique strengths of the Japanese industry?

In my opinion, there are two kinds of manufacturing; one is a modular architecture which is a combination of different modules. In this type of manufacturing, the components are already available in the market and firms simply combine various combinations of components to complete and assemble the final goods. The other type of manufacturing relies on specialized and specific components that vary from model to model. These two philosophies are very different. This second type of manufacturing; the “model-by-model optimization” philosophy is what Japanese firms really excel at and it works well for products such as automobiles as well as machinery. The “modular combination” philosophy works well for consumer electronics, but this area as we have all seen currently is dominated by Chinese firms who have leared quickly. Japan has simply lost its position in this area. i-PRO should be on this side of manufacturing, the modular combination philosophy, but before we spun off from Panasonic, there was a period of time where we utilized model-by-model optimization.

Panasonic’s philosophy was to design model by model, and it took them a long time to develop products. The number of models they could introduce in a year was so limited. If you take 2019 as an example, the number of new products introduced was less than 20. If I can remember correctly, it was around 13-19 because it depends on how we define separate models. Last year alone, i-PRO introduced 90 models because we changed the design policy.

Electronics really needs the taste of the modular combination philosophy, but if we choose on this, we have to compete with China, Korea, and Taiwan, and honestly speaking, our odds of winning are dropping less and less. i-PRO is now pursuing a hybrid of both approaches. (a hybrid approach, both model-by-model optimization and modular combination.) To win this competition, the key word is time.

i-PRO engineers must learn two “Chains.” One is Engineering Chain, and the other is Supply Chain. The Engineering Chain refers to the process of new product development. This chain starts from market analysis, product planning, engineering design, prototyping, to preparations for mass production. The supply chain starts from demand forecasting, sales pitch, quotation, raw material procurement, order receiving, manufacturing, warehousing, delivery to cash collection. Engineering chain and supply chain need to be connected very tightly, and a good example of this might be Apple. In California, Apple has its engineering, but they also have their manufacturing in China. I think Apple is operating these chains productively, but not many companies can operate these chains as effectively as Apple if the locations are widespread such as in this case.

We use mass production to produce multi-purpose modules, but the actual final assembly is based on the purchase order. Basically, I like to think of it as an on-demand assembly system here in Japan. We have our factory in Saga, Kyushu and our design laboratory is located in Fukuoka, only a one-hour drive, meaning communication is quick. That is why we were able to introduce over 90 new models last year alone. The overriding objective is to link Engineering Chain and Supply Chain tightly thus shortening the lead times of both new product development and order-to-delivery, and in that respect, modular design is indispensable. The manufacturing side and the R&D side must be connected as one team to create this modular setup.  


Can you speak on how you implemented this new business process you have talked about?

Designing a multi-purpose module is actually difficult, and that is because the modules must have some redundancy and allowance for combinations. When I changed the policy from the model-by-model optimization policy, there were a lot of discussions among the engineers, and they had a lot of objections. Finally, they understood my thoughts and agreed to start designing multi-purpose modules but that was a year-long struggle for us to overcome. Now we have a number of multi-purpose modules which can be applied to many model variations. High-mix-low-volume is the key to modern-day Japanese manufacturing with the addition of short lead times. Everything must be connected digitally, from demand forecast to material procurement, all the way to customer feedback upon delivery.

We know many AI vendors must work onsite at their client company to develop solutions through trial and error. Furthermore, when it comes to camera vendors, they have not coped well enough to handle AI because there are few makers in the world that can manufacture cameras in high-mix-low-volume with prompt delivery. In saying that, 60% of data used with AI is images, which means that cameras play an essential role in working with it, with the potential demand in the AI camera market for industrial use alone estimated to be worth around JPY 400 billion. Can you explain your vision when it comes to cameras relating to AI? As a company that is focusing on high-mix-low-volume production, what opportunities are there for AI?

AI is diverse and there really are so many different use cases and user experiences. As each experience is different, so should the AI application be different. The industry is quite fragmented and there is no one big AI solution. AI vendors have to be affiliated with their users and because of this fragmentation, right now we are not interested in becoming an AI solution provider.

We will provide AI tools to many AI providers. I want you to think about a fashion photographer for a second. When they are on the job, they carry many different cameras and lenses. Why do they do this? The reason is that each camera and lens offer a different feature, and combination of different bodies and lenses should produce an optimum output. All cases are different and based on that, they must change the tools they use for the job. Security is the same, and I think that AI for industrial cases is the same too. AI needs the best imaging equipment but thanks to the emergence of quality cameras built into smartphones, the variations of cameras are decreasing. 30 years ago, when I was young, there were many camera makers, so the lens variations were enormous. Nowadays people have quick, cheap, and easy access to quality cameras on a smartphone. Unfortunately, not everyone is a world-class photographer, and not all smartphones fit a diverse photographic environment, often photos taken with devices like smartphones can't be good enough for AI. Therefore, AI developers need to recover the poor image by developing good software. AI likes good images and if the quality of the image is good, the software burden is low. We as a company would like to provide many different cameras to fit many different cases, but as a manufacturer, we cannot develop too many models at the same time. That is why we employ modular design to account for different combinations.


Do you believe that AI has not been widespread because good cameras are still unavailable?

Yes, and in 10-20 years, if good hardware is provided, I think AI applications will be more diverse. When you talk about costs for AI, that cost incurs during system integration. If the software burden is reduced by introducing good hardware and trial and error duration is minimized, I think AI availability will be much increased.


You have defined yourself as a hardware manufacturer, and as such collaborations are important. Are you looking for any partnerships in overseas markets?

Yes, of course. Many of our partners right now are outside of Japan, especially in the US and European markets. We are continuously looking for partners and AI vendors. In fact, those system integrators and AI vendors are happy now that they know we will never be their enemy and that is because we abandoned the end-to-end solution business. We do not want to compete with system integrators or AI vendors, and now we have a healthy relationship where they trust us.

From Panasonic’s point of view, they provide end-to-end solutions, from software to hardware, meaning that sometimes they were competing with our customers. The first question I asked our employees when I came to i-PRO was, “Who are our customers?” What I told our R&D department was that system integrators do not make cameras. They buy cameras from outside. We should not be competing with potential customers, and instead what we need to do is sell good cameras to all of those system integrators. It was my mission to first identify exactly what company i-PRO was. I wanted to define i-PRO not to compete with our potential customers, system integrators, who provide end-to-end solutions.


One camera that piqued our interest is the Multi-Directional PTZ line of cameras with specific products such as the WV-X86530-Z2. It is the smallest and lightest outdoor multi-pan-tilt-zoom (PTZ) camera of its kind in the industry. This camera utilizes AI analytics for multisensory on the PTZ. What motivated you to develop this specific product?

Actually, we were behind and we are not the first company to introduce this kind of product. We felt that we should not make the same thing and thought about the uniqueness we could provide. I think the key to our product's uniqueness is in the Edge AI capability. Additionally, the WV-X86530-Z2 is a small size and is quite beautiful aesthetically. That small size enables us to lower costs too, which are passed to the end users. In the case of this product line, size was our first priority. The customer can choose what AI components can be installed from our menu. One example might be human detection, so we have developed standard application venues so customers can choose from pre-existing application optimization, or alternatively, they can develop their own applications. These applications are directly connected to the large VMSs. We actually introduced a plugin called Active Guard, which is directly connected to these AI applications acting as a software analytics engine. For example, we might set it to search for people wearing red shirts between the times of 12:00 and 13:00. These parameters can be set with a few clicks. This plugin works on Genetec and Milestone, global major VMSs in addition to Video Insight, i-PRO’s original VMS. We are open. Our edge-AI works with most major platforms. In this respect, we believe we are the most advanced.    

A UN report in 2022 warned that people’s rights to privacy are coming under greater pressure, but conversely, there is the opinion that we need more security in order to combat crime and terrorism. Where do you stand on this debate?

There are frequent talks among CEOs in the industry and in particular, I had a lot of opportunities to talk to some key players during ISC-West 2023 in Las Vegas this year. Trust is an overriding concern for us, but trust is not made in a day. We need to build it piece by piece and that needs to start with our corporate belief or mindset. To do that, we first made a code of conduct, which outlines our contributions to society and our respect for basic human rights. We believe very strongly that we have to protect human rights. The next step was to decide what rules we need to comply with because as you know, rules and regulations vary from country to country.

For us, the most important market is obviously the North American market, so National Defense Authorization Act (NDAA) compliance is crucial. During the time when we were a part of Panasonic, we were not NDAA compliant, and that came because Panasonic had such strong ties with China. There were many sensitive issues in NDAA discussions at that time. However, i-PRO is a company of security products, so a declaration of NDAA compliance is key, otherwise, we cannot survive. We decided to eliminate all the Chinese chipsets and that was only possible because we gained our independence from Panasonic. In addition to NDAA compliance, cyber security is the utmost concern.  We are actually now level three of FIPS 140-2 in our security standards, which by the way, not many companies are. At level three, we must use expensive materials.

In fact, cost and trust are always trade-offs. For example, we had intense discussions on which chipset we should use for our new recorders. There are many chipsets as you might know already, chipsets from Taiwan, China, and Korea. From the cost perspective, Taiwan is the best. From a security perspective, as you can guess, China is not great, and even Taiwan is in a gray area. Finally, we decided to use Intel for our new recorders. At the time, there were many objections and many thought that we should use the cheaper. Obviously, going with Intel meant that our costs became much higher than our competitors. But we finally chose Trust.

To sum up, basically, once we have chosen the rules and regulations that we will comply with, then we move on to align all business processes such as, procurement, technology, architecture, partnering, and manufacturing with the compliance decision. This is the framework where we can build trust among our partners and it is the most important aspect for survival in this industry. Having said that however, it is not easy, but I think having our independence now means that the overall picture is a little clearer for us. We can focus on the security industry and have a clear picture of what we are aiming to become.


In addition to the security sector, we also know that you make cameras for industrial and medical use. You have a 30-year history developing cameras for medical devices during your time with Panasonic. Medical cameras account for about 15% of your total sales and as you know, they have different requirements and aspects that need to be met. Color reproduction is one and the needs vary widely from those of security cameras. What are some of the anticipated challenges you have to overcome as you continue to grow this medical vision aspect of your business?

During our time with Panasonic, medical cameras and security cameras were handled by separate teams. Although we used the same technology, clearly there was a divide and no synergy between the teams. Last year, we merged the technology base between medical and security. There are many opportunities now to enjoy the synergy between medical and security, and one example might be the chipset. For medical devices, we use a field-programmable gate array (FPGA) whereas for security, we use SOC. Even though they are different, there are some rooms to merge these technology areas. If we can merge the common areas of technology, we can reduce costs, reduce development times, and enhance the variety of products. There was no modular concept with medical cameras because the medical business is based upon R&D contracts with clients. Our salespeople interview the customers for their needs, and engineers make it happen through the model-by-model optimization approach. What this means is that the finished goods cannot be sold to another customer.

Last year we changed this, introducing our modular design concept to the medical field. We are preparing a medical camera catalog, which by the way did not exist in the past. We hope to have the first version ready by the summer. With this catalog, we can approach small and medium-sized medical equipment manufacturers. There are so many small and mid-sized medical equipment factories globally because the regulations vary from country to country. To cover such a large number of potential customers, we will work to show off our medical vision catalog. I want our company to reach a point where our clients  search the catalog for their application, click, and by the next day they have our cameras at their factory or location ready to install. I would like to apply the same concept to industrial cameras also.

This technology is something that comes from our heritage with Panasonic and the basic technology applied to security and medical are essentially the same. There are so many areas beyond security and medical that we can use our technology in. To do so, however, we need capital, a lineup, and a way for potential customers to see what products we are offering.


i-PRO has different operations in the US, Canada, China, Europe, Asia Pacific, and Singapore, and earlier you highlighted that the most important region for you is North America. Are there any other regions or countries you have identified for growth in the mid-term?

That is a sensitive question to answer and unfortunately, the global market is divided. In particular, for the security industry, the world is divided into two; one side is America and Europe, the liberal economy, and the other is China, the controlled economy. Obviously, for our business, we are now focusing on the West. We are aware that there is a big market in China, but despite that, our focus will remain on Western, high-end clients, with businesses where quality and trust are the most important elements. As I mentioned earlier, we are the kind of company that is not afraid of spending our time and money on improving reliability and quality. Of course, the costs are high but there are so many potential customers. In each region, there is a certain market segment  that does not care too much about security and will go with cheap solutions. We do not intend to compete in  such a sector and we plan to continue to focus on the high-end market.


Imagine that we come back and have this interview all over again on the last day of your presidency: what goals would you like to have achieved by then?

Some people may say that i-PRO is a Japanese company, but I do not think that is the case. Sure, the origin is Japan, but I truly believe that the company has become a multinational company and I am proud of that fact. I want to reach the point where people actually forget that i-PRO’s origin was Japan and really consider the company to be a global one.

i-PRO is open. We must be a trusted partner among AI professionals of security, medical, and industry all over the world. By trusted, I mean our Engineering Chain and Supply Chain enable the latest technology to be transformed into edge-AI cameras faster than any other else. We will deliver exactly what is demanded just in time as they are demanded. It is i-PRO’s operational excellence. By this, we can help all AI engineers develop AI applications easier and faster so that more AI applications help our society. I wish that 5 to 10 years later, i-PRO’s AI eyes are everywhere to watch for security, save lives from difficult diseases, and enhance productivity of all industries.   

Panasonic Spin-Off Camera Manufacturer’s Growth Strategy

Interview conducted by Karune Walker & Paul Mannion