Implementation of PoC (Proof of Concept) is indispensable for launching a new business. In the past, PoC was mainly used by pharmaceutical companies and IT companies, but in recent years, it has spread to a wide range of industries as a touchstone for measuring the success or failure of business projects, and many companies are working on it. We will explain in detail the meaning of PoC, how to do it, points to avoid failure and case studies for those in charge of companies considering implementing PoC.
Table of contents
What is PoC (Proof of Concept)?
Definition of PoC
POC (Proof of Concept, reading: Pock or POC) is to confirm whether a new technology, theory, principle, method, idea, etc. It is an experimental verification process.
In Japanese, it is often translated as proof of concept, and in some cases, it aims to identify the necessary elements and specifications to obtain the desired effect. It can be said that the characteristic of PoC is that it is not a verification based on theory or calculation, but a simple version of a product or system is created, and concrete verification is performed by actually using it.
Advantages of PoC and reason for implementation
Can reduce risk
Because PoC is conducted on a small scale and a trial basis, it is possible to reduce product and system development risks. Although the scale is small, it is possible to verify the operational stability and usability of the actual product and measure whether it is technically feasible.
Also, in the process of making the actual product for PoC, it may be predicted that the development process will be prolonged. A representative advantage of PoC is that it reduces development risks by assessing feasibility from multiple perspectives.
Can reduce costs and man-hours
Another advantage of PoC is that it can reduce costs and man-hours.
Suppose you are trying to come up with a new idea. The idea may be wonderful, but whether it can be realized, and even if it is realized, whether it will be accepted by society and used is another matter. As a result, if the project proceeds based on assumptions without verification, and the feasibility is low as a result, the costs and man-hours that have been spent up to that point will be wasted.
However, PoC makes it possible to judge whether it is feasible at an early stage and to easily consider the subsequent development, so it is possible to reduce costs and man-hours.
A powerful tool for making decisions on investment and technology development
If the PoC proves that there are no technical problems and that the product or service is effective for its purpose, the possibility of realizing the product or service will be greatly increased, so investors will also gather investment in the project. easier.
Investors are reluctant to spend money on things that have no visible effect and are difficult to realize. To avoid loss risk, the results of PoC will be a powerful decision material.
Also, if sufficient effects and results are shown at the PoC stage, the developers will be able to decide that further development costs and time are unnecessary.
It can be said that PoC will be a powerful decision-making material for many people involved in the project.
Industries using PoC
The pharmaceutical industry has a long history of using PoC. New drug research and development and PoC are inextricably linked.
New drugs can be mass-produced only after they have been administered to animals or humans and are useful and effective. In general, the development of a new drug takes a long period, from 9 to 17 years, to conduct verification.
According to the Pharmaceutical Society of Japan, new drugs are approved in three stages. When it is received, it is possible to proceed to the final stage as “PoC has been obtained”.
PoC is also used in the movie industry. Since production costs for feature-length films such as 90 minutes and 120 minutes are very expensive, there are cases where short films or short videos are first created and then used for PoC.
By creating a short film that incorporates ideas, stories, new technology CG, etc. at a low cost before creating a feature film, we can see the audience’s reaction and the audience’s expectations for the movie to be produced in the future. can increase
In addition, the creators can also find issues and checkpoints for improvement when creating feature films.
In the IT industry, PoC is performed when introducing a new system or building security.
For example, when proceeding with the introduction of a system to a client, we make a hypothesis, create a model type based on that hypothesis, conduct a PoC at the actual site, and verify the effectiveness of the system by verifying the hypothesis. Then, we will look at feasibility such as whether the new system can be introduced and whether it can be used in business after the introduction.
Since it costs money to develop a complete system, the industry often conducts demonstrations with the minimum functions that match the purpose.
Research and Development
In the field of research and development, PoC is mainly conducted for budget acquisition. As you know, research and development are very expensive. By conducting a PoC and demonstrating the feasibility, it is possible to gain approval and obtain a budget.
In cutting-edge fields such as biotechnology, there are many cases in which PoC demonstrations and proposals are made to move toward realization.
Differences between PoC, proof-of-concept, and prototype
Difference from demonstration experiment
As mentioned earlier, PoC is often translated as proof of concept, but it is sometimes translated as proof of concept, and the current situation is that it cannot be said that there is a clear line.
On the other hand, while PoC is conducted to verify the feasibility to achieve its goals, there is also the view that the purpose of demonstration experiments is slightly different because it verifies problems.
However, even in PoC, new challenges may be found through feasibility verification. Therefore, PoC and demonstration experiments can be considered almost synonymous.
Difference from prototype
While POC refers to the process of verifying ideas and technical feasibility, a prototype refers to the process of creating a prototype after the direction of the idea and technical feasibility have been determined.
In other words, the PoC is work to find a goal, and the prototype is the final confirmation of something that already has a clear goal. It is easy to confuse both because they are both prototypes, but there is a clear difference in purpose.
Therefore, as a flow, it is often the case that the feasibility is first confirmed by PoC, and then the prototype is produced.
PoC in DX
PoC is the key to success in DX (Digital Transformation) as well. DX is a concept that “IT will change people’s lives for the better in every aspect by taking root in the foundation of people’s lives.” Mr. Sudo of Kaizen Platform says that
DX promotion is “small” and “fine”.
The key to the success of DX is to first implement a PoC and produce results in an easy-to-understand form such as numerical values, verify the value of the project, and promote it while involving other departments that recognize its value. It is said that
And the PoC must be “small” and “detailed”. This is because if you try to do everything at once, there is a high possibility that the project will fail due to many obstacles such as cooperation with uncooperative other departments who do not understand the value.
Accumulating prototyping for demonstration and verification, even if it is small, will lead to the final goal of the project.
PoC method, 4 steps
STEP 1: Clarify the purpose and goal of PoC
PoCs need to be decided in detail before starting, such as why you are doing it, what effects and data you want to get, and what data you need. This is because if the purpose and goals are not clear, the guidelines will be blurred and the PoC will not be able to be used effectively. Also, the PoC itself may become the purpose or goal.
Let’s assume “introduction of RPA and AI for routine work efficiency”. In this case, it would be a good idea to set specific targets, such as how much effect can be expected from reducing work hours, whether the effect will differ depending on the department or period, and the behavior after achievement.
STEP 2: Determine specific implementation details such as verification methods
Once you have decided on a clear purpose and goal, the next step is to decide on the implementation details.
Here too, it is important to be specific about verification methods. By clarifying the purpose of the PoC, you should be able to see what you will need. Therefore, it is better to create the minimum required to obtain the necessary data.
The important thing at this time is to be conscious of creating something that is as close to the actual site as possible from the user’s point of view. If you can create something from the user’s point of view, you will get more effective and concrete results.
In the example of “Introduction of RPA and AI to improve the efficiency of routine work” mentioned above, from the perspective of the person actually in charge of the work, what kind of system should be used to automate what kind of work? It is necessary to decide whether to measure results. If you see their site before implementing PoC, it will become more specific from the user’s point of view.
STEP 3: Demonstrate
Once you’re ready, it’s time to prove it. We will try and prove what we have prepared on the real site. Even if it is implemented on a small scale, it will be possible to demonstrate with higher objectivity and accuracy by having all the subjects use it and evaluate it.
Even in the example of “Introduction of RPA and AI to improve the efficiency of routine work” mentioned above, evaluations and data obtained from only a few people will be biased, and so have all the departments that have introduced them use it. By doing so, you will be able to obtain more objective and persuasive data.
STEP 4: Evaluate PoC results
The final step is to evaluate the results based on the data obtained from the PoC.
It is a good idea to keep in mind that the evaluation here needs to be strictly evaluated without being overwhelmed by emotions. Using the data as a reference, we will verify practicality, risks, and evaluations from investors.
If the results are positive, the project will move forward toward full implementation or commercialization.
However, since PoC is working to confirm the feasibility, it is fully possible to expect negative results from time to time. Even at that time, the issues and problems that were discovered for the first time after the verification were visualized and can be utilized in the next verification.
“Introduction of RPA and AI to improve the efficiency of routine work” may have yielded more results than expected in STEP 1, and the evaluation at the site may have been excellent. Alternatively, new work may be required to implement the system, resulting in far less efficiency.
In any case, it is important to take the voice of the real site seriously. There is no change in the fact that the PoC provided us with useful materials that led to various decisions.
4 points for not failing in PoC
Start small and fast
The iron rule is that PoCs must be small and start quickly.
If implemented on a large scale, in addition to being costly, it will inevitably take time, and there is a possibility that it may be unclear what the purpose was. Therefore, it is important to narrow the scope of the purpose and accumulate small contents.
When doing PoC, it would be good to be aware of a small start.
Validate under the same conditions
It is also important to verify the PoC under the same conditions as the environment in which the system is intended to be introduced.
If the conditions are the same, more accurate data can be obtained, and subsequent decisions can be made easier.
If you get a positive result, it will be more persuasive to other departments and investors.
If you cannot prepare the same conditions, you can obtain useful data by creating a virtual similar environment in another place.
It can be said that it is very important to verify under the same conditions to use PoC more effectively.
Learning from failures through the PDCA cycle
It is not a mistake to have issues and problems by doing PoC.
This is because the issues and problems are only discovered after verification, and cannot be recognized in the first place unless PoC is performed.
What is important is to think about the issues and problems identified by the PoC, why they happened, and what to do next to get better results, and then run the PDCA cycle.
As the saying goes, “Failure is the mother of success.” PoC can be done with less risk. The data obtained from failures will also be effective if used next time.
Not intended for PoC
If PoC is performed continuously, the purpose of PoC may become the purpose and the original purpose and goal may fade away.
Therefore, it is important to keep in mind that PoC is a means, not an end.
The purpose is to judge whether the product or the system to be introduced is feasible, and implementing PoC is only a means to achieve the purpose.
If you get lost, it would be better to organize once again why you are doing this post. Validation without purpose is meaningless.
PoC, 10 examples of demonstration experiments
BOLDLY (former SB Drive): Practical application of self-driving buses
BOLDLY conducts demonstration experiments at least once a month, and conducted 23 times in 2019. We are conducting demonstration tests mainly on fixed-route buses, in theme parks, and at events, with a total of 5,413 test riders and a driving distance of 12,019km.
As of June 2020, we are also conducting test runs on public roads for practical use, and have succeeded in running a vehicle designed on the premise of automatic driving on public roads without creating an enclosed space for the first time in Japan.
Taisei Corporation: Construction site safety management using 5G “i-Construction”
“i-Construction” promoted by the Ministry of Land, Infrastructure, Transport and Tourism refer to an initiative aimed at improving worker safety management and productivity by introducing it to construction sites.
Taisei Corporation conducted a demonstration experiment using 5G at a tunnel construction site to realize i-Construction. In recent years, fatal accidents have occurred due to cave-ins and gas spouts in tunnel construction, and the safety of workers must be given top priority.
By installing a sensor that uses 5G communication, it is possible to monitor carbon dioxide in real-time and notify workers in the event of an emergency.
In addition, we are verifying the feasibility of remote control of construction machinery assuming disasters, and have succeeded in remote control from a point about 1,400m away. It is expected that safety confirmation at disaster sites can be confirmed by remote-controlled construction machinery.
Softbank and WPC: Truck platooning using 5G
Softbank and Wireless City Planning have succeeded in platooning trucks using 5G for the first time in the world. The demonstration experiment was conducted on a highway that is always used by long-distance drivers.
In February 2020, a demonstration experiment was successfully conducted by platooning of three trucks traveling 20km on the Tomei Expressway at a speed of about 80km/h.
Using 5G communication, the distance between the leading manned vehicle and the trailing unmanned vehicle is controlled to 10m and is expected to stabilize control between vehicles and improve visibility when confirming safety behind. In addition, the result is expected to solve the shortage of drivers due to the declining birthrate.
Ministry of Agriculture, Forestry, and Fisheries: Smart Agriculture Demonstration Project
The Ministry of Agriculture, Forestry and Fisheries is implementing a “smart agriculture project” to resolve the labor shortage in agriculture. We are conducting demonstrations in all kinds of crops such as fields, fruit trees, and livestock all over the country.
At Shinjuku Seika in Miyazaki Prefecture, employees used to enter cultivation information from their mobile phones, but the use of ICT has reduced the amount of time it takes to enter data, reducing the work hours per worker by 50 hours per month. It says.
In addition, Moriyamaen Co., Ltd. in Aomori Prefecture used a cloud-supported application to manage the production of apples, reviewing management work and improving work efficiency.
It can be said that these are actual examples of PoC that have demonstrated the efficiency of farm work.
National Research Institute for Earth Science and Disaster Resilience: Disaster prevention chatbot “SOCDA”
The National Research Institute for Earth Science and Disaster Resilience is conducting a demonstration experiment of the disaster prevention chatbot “SOCDA” to support the decision-making of disaster response organizations and the evacuation of citizens in the event of a disaster. In the event of a disaster, the Twitter tweets of citizens are visualized on a map, the damage is analyzed by the system, and the safety is confirmed.
Demonstration experiments are being conducted in Itami City’s flood prevention map training (training to consider flood damage countermeasures using maps), and the aim is to automate the entire process. It can be said that when SOCDA becomes practical, it can be expected to reduce disasters.
Pixar: Demonstration short film
Pixar is making a demonstration short film. The main aim is to incorporate the new technology of CG into short films and gain customer appreciation.
The blockbuster movie “Toy Story 2” incorporates animation that makes the texture of cloth and facial expressions clearer and smoother. It is said that these technologies were implemented in a short film produced before the release of Toy Story 2.
PoC is active in the movie industry, and Pixar’s case is the best example. It will be interesting to see what technology is used in future Pixar short films.
Toyota: Smart City “Woven City”
In 2021, Toyota plans to set up a smart city called Woven City in Susono City, Shizuoka Prefecture, which will serve as a model case for cities shortly.
In this city, we plan to demonstrate by actually using all kinds of cutting-edge technologies such as CASE, which refers to next-generation automobile technologies and services such as connected cars, autonomous driving, electric vehicles, AI, personal mobility, and robots, in the city. It is said that
There are three distinct roads. There are roads for vehicles, roads for both pedestrians and vehicles, and roads for pedestrians only, and on the roads for vehicles only, a fully automated car called e-Palette will run.
In addition, on roads where pedestrians and vehicles run together, slow-speed personal mobility coexists with people.
The aim is to create new value and business models and build a “town platform” through a demonstration city that conducts demonstrations throughout the city.
Itoki: Demonstration experiment office “ITOKI TOKYO XORK”
Itoki built “ITOKI TOKYO XORK” as a demonstration experiment when introducing 5G to the office as a joint research of six parties. The purpose here is to confirm what devices will be used when 5G is introduced and how the way of working will change.
There are 93 rooms in the office, which is quite a lot considering the number of 800 employees. In addition, new technology has been introduced into the system. That is a teleconferencing system using MR. MR can be said to be an evolved version of the current video conference system, as it gives the feeling that the person in the remote place is right in front of you.
In the future, this office will carry out demonstration experiments as needed to encourage work style reforms in the three areas of “workplace,” “IT,” and “behavior.”
Fujitsu: Using IoT to visualize the flow of people in tourist spots
Fujitsu conducted a demonstration experiment of a system that visualizes the flow of people using IoT in three cities and six towns and villages in Hokkaido. Using Wi-Fi packet sensors installed in the city, we anonymously collect unique IDs for smartphones from 40 Wi-Fi packet sensors in the city. Analyzed the congestion situation, movement routes, etc.
Based on the data obtained from this demonstration experiment, it is said that it will be used for tourism measures based on needs in the future.
Toshiba Tec: Unmanned stores using AI and IoT
Toshiba Tec conducted a demonstration experiment of an unmanned store using AI and IoT at the “Oval Court Osaki Mark East Office Shop”. The target of the demonstration experiment was Toshiba Tec’s employees.
As the name suggests, unmanned stores do not have staff to serve customers, and the stores are operated using smartphones, tablets, and sensor technology.
A distinctive feature is the reflectionless use of a sensor camera. Reflectionless means eliminating the hassle of paying when purchasing a product.
A sensor camera identifies the product and the user, allowing the user to take home what they want without going through the cash register. Payment is deducted from the employee’s salary.
In the future, we will aim for practical applications to solve the shortage of human resources due to the declining birthrate and to realize a comfortable shopping environment for users.
PoCs are increasingly being used in a variety of ways, such as demonstrating innovative technologies, introducing new products and services in-house, and gaining support from investors.
As mentioned earlier, PoC and demonstration experiments daily for new technologies such as 5G, AI, and IoT. In particular, in promoting DX, it is important to carry out “small” and “fine” demonstrations and verifications many times in PoC, and how to proceed to lead the project to the goal.
PoC will become the standard for companies in the future. We would like you to use PoC to develop society through the development and introduction of efficient products and services that reduce risk.