How to Conduct a Hardware Proof of Concept (PoC): 5 Steps, Costs, Duration, and Failure Cases

Narrow down by specifying conditions

現在2180件がヒットしています。check

Basic sensor board module wireless

PoC (Proof of Concept) Edition: Verifying if the Proof of Concept "Works"

This article is the first in a three-part series on proof-of-concept (PoC) for hardware commercialization. Hardware commercialization progresses through three stages: PoC → PoV → PoB, gradually increasing the level of verification.

Part 1: Proof of Concept (PoC) - Checking if it "works" (This article)
Part 2: Proof of Value (PoV) - Verifying whether it's useful and worthwhile (Under preparation)
Third installment of PoB: Testing whether a business can be sustained (in preparation)

step	question	Main deliverables	Estimated duration
PoC (Proof of Concept)	Is it technically feasible? = Does it work?	Working prototype + decision report	2-6 months
PoV (Proof of Value)	Is it useful in the field? Is it worth the investment?	Report on measuring the effectiveness of small-scale implementations	3-6 months
PoB (Proof of Business)	Will it be sustainable as a business?	Mass production design + business plan + market launch preparation	6 months and up

Hardware Proof of Concept (PoC) and software Proof of Concept (PoC) are different things.

If you search for "how to proceed with a PoC," you'll mostly find articles about software and AI. However, unlike software, where you can simply rewrite the code, hardware requires redesigning the circuit board, resourcing components, and remanufacturing the casing. If you proceed without understanding this difference, the PoC itself will not be successful.

The characteristics of the hardware PoC (Proof of Concept) are as follows:

A proof-of-concept experiment to verify whether ideas for electronic devices and IoT devices are technically feasible.
The key feature is that it involves actually creating and testing "working objects."
Unlike software, redesigning, sourcing parts, and manufacturing the enclosure requires several weeks to several months for each revision.

Please also refer to our related articles: How to Choose the Right Outsourcing Partner for Product Development to Avoid Failure | Differences between OEM, EMS, and ODM, 5 Patterns of Failure in Outsourcing Electronic Device Development, and [Essential Guide] Wearable Development | Requirements by Application and 3 Implementation Challenges.

[Comparison Table] Software PoC vs Hardware PoC

Comparison item	Software Proof of Concept (PoC)	Hardware Proof of Concept (PoC)
Subject to verification	Algorithms, UI/UX, data processing	Sensor accuracy, communication stability, housing durability, battery life
Things you need	PC, cloud environment	Circuit boards, components, sensors, housings, power supplies, communication modules
Speed of redo	Code fix and redeployment within a few days	Redesigning the circuit board, procuring components, and then assembling it takes several weeks to several months.
Verification environment	The entire process can be completed on a development PC or in the cloud.	Field testing in factories, outdoors, stores, etc. is essential.
Cost structure	Development man-hours are the main factor.	The cost of parts, circuit board manufacturing, and enclosure manufacturing are added on top of each other.
Transition to mass production	Switch deployment target	A complete overhaul of the mass production design is necessary.

If you proceed with the same mindset as a software proof-of-concept (PoC), you'll end up with situations like, "We just built it, but it didn't serve as a valid test," or "It worked in the development lab, but it didn't work in the field."

The proof-of-concept (PoC) process doesn't end in one go; it involves three stages: PoC → PoV → PoB.

Commercializing hardware isn't just about confirming whether it "works." The verification process progresses in the following order: "Does it work (PoC)" → "Is it useful and cost-effective (PoV)" → "Is it viable as a business (PoB)."

There are three stages: PoC → PoV → PoB. The level of verification progresses from "Does it work? → Is it useful/worth it? → Is it viable as a business?"

[Comparison Table] Three Stages of Verification

step	question	The main thing to verify	Main deliverables	Estimated duration
PoC (Proof of Concept)	Is it technically feasible? = Does it work?	Sensor accuracy, communication stability, and basic operation.	Working prototype + Go/No-Go decision report	2-6 months
PoV (Proof of Value)	Is it useful in the field? Is it worth the investment?	Effectiveness in real-world environments, ROI, operational burden, and user acceptance.	Report on measuring the effectiveness of small-scale implementations	3-6 months
PoB (Proof of Business)	Will it be sustainable as a business?	Mass production costs, sales channels, revenue structure, and support system	Mass production design + business plan + market launch preparation	6 months and up

The most important thing to remember is not to jump straight into mass production design based solely on the "Go" from the Proof of Concept (PoC). "Technically working" and "creating value in the field" and "being profitable as a business" are two different things. Skipping the PoV will result in mass production of a product that "works but nobody uses/is not cost-effective."

Overall outline of the commercialization schedule

The roadmap will vary depending on the complexity of the product and whether or not certifications are required, but the table below shows a two-year roadmap as an example. It outlines three stages on a timeline, clarifying "what tasks are performed at each stage and how long it takes."

Commercialization roadmap. Includes the PoC/PoV/PoB process and Go/No-Go milestones (①②).

Timing (approximate)	step	What to do in this process
Months 1-3	PoC: Preliminary Evaluation	Use an evaluation board (EVM) to identify potential candidates, select technologies, and set KPIs and decision criteria.
Months 3-6	PoC: Basic Evaluation	Prototype creation, lab-based operational verification, Go/No-Go ①
5-9 months	Proof of Value: Value Verification	Small-scale implementation in a real environment, effectiveness measurement, ROI estimation, identification of operational challenges, Go/No-Go ②
8-14 months	PoV → PoB: Realizing the business model	Realizing products/services and revenue structures, and verifying sales channels and methods.
10-18 months	PoB: Practical/Mass Production Design	Finalizing mass production specifications, obtaining certification, and creating mass production prototypes.
12-24 months	PoB: Pilot program → Market launch	Pilot testing at partner companies, initial lot quality assurance, and preparation for mass production and market launch.

For example, starting a small-scale proof-of-value (PoV) test immediately after the basic evaluation and running it partially in parallel with the mass production design will reduce rework. If everything is done in series, the whole process will take about 1.5 times longer. Optimizing the order and timing will be discussed in more detail later in the section "Is the order of your company's commercialization schedule correct?".

How to Conduct a Hardware Proof of Concept (PoC): 5 Steps

From here, we will take a detailed look at the first of the three stages—the PoC (Proof of Concept)—in five steps.

The five steps of a Proof of Concept (PoC): (1) Hypothesis setting → (2) Technology selection → (3) Prototype creation → (4) Field testing → (5) Go/No-Go decision

[Comparison Table] 5 Steps and Deliverables of PoC

step	Things to do	Artifact	Estimated duration
① Setting a hypothesis	Define verification items, KPIs, and judgment criteria.	PoC design document	1-2 weeks
②Technology selection	Select sensor type, communication method, and microcontroller.	Technology selection table, evaluation board (EVM) verification results	2-4 weeks
③ Prototype Production	PCB design, component mounting, enclosure fabrication	Working prototypes (1 to several units)	1-3 months
④ Field test	Verified in actual usage environments.	Test Report	2-4 weeks
⑤ Go/No-Go decision	We will compare it with KPIs and decide whether to proceed with mass production.	Decision report + action plan	1 week

① Hypothesis setting: First, decide "what to verify."

The worst thing you can do in a Proof of Concept (PoC) is to "just try to make something that works." A PoC conducted without a hypothesis becomes ambiguous in what constitutes success, and ends up being an unfinished project.
The format for setting hypotheses is as follows:

What we want to verify: "Can we measure with the accuracy of □□ under the environment of △△ using the 〇〇 sensor?"
• KPIs: "Measurement accuracy within ±○mm", "Communication success rate of ○% or higher", "Battery life of ○ days or more"
Criteria for decision-making: "KPI achieved → Go" "Partially unmet → Revise hypothesis and conduct another PoC (Pivot)" "Technically difficult → No-Go"

Even if the final decision is a "no-go," it's not a failure. There's great value in being able to withdraw at the proof-of-concept (PoC) stage, rather than discovering it doesn't work after mass production.

② Technology selection: "Get a rough idea" using the evaluation board.

Instead of immediately designing a circuit board, we perform preliminary verification using an evaluation board (EVM) provided by the sensor manufacturer. This allows us to verify whether the sensor will achieve the desired accuracy at low cost and in a short period of time, without the need for soldering.

However, there are pitfalls in simply applying the settings from the evaluation board directly to the actual device. In one wearable device project, applying the gain values used on the evaluation board directly to the actual device resulted in poor performance, requiring readjustment. Because the circuit configuration differs between the evaluation board and the actual device, re-verification of the settings is essential.

③ Prototype creation: The goal is for it to "work," appearance is not a concern.

The purpose of the prototype is to verify functionality, so a high level of aesthetic perfection is not necessary. A 3D-printed casing and a hand-soldered circuit board are sufficient. One customer explicitly stated, "These five units are for proof-of-concept, so only internal staff will use them. It's okay if they're not pretty, just get on with it."

Macnica 's PoC development team has a track record of completing prototypes in as little as one month. The ability to quickly create a prototype using readily available parts and show a working product to superiors and stakeholders to get their feedback is crucial for internal decision-making.

Handmade sensors for food and beverage digital transformation

Homemade Sensors for Environmental Sensing

Mockups made with a 3D printer

④ Field testing: Testing outside the development lab.

This is the most often overlooked step in hardware proof-of-concept (PoC). A prototype that works in the development lab may not work the same way in a real-world usage environment.

In one IoT sensor project for a particular piece of equipment, two prototype units delivered malfunctioned in the field. The causes were a failure in the LTE module's response and a broken power supply line that occurred during shipping and packaging. Problems that would not occur in the development lab occurred at three stages: transportation, installation, and the outdoor environment.

environment	Risks to be examined
plant	Vibration, dust, electromagnetic noise
outdoors	Temperature changes, humidity, direct sunlight
Restaurant	Water droplets, oil fumes, WiFi congestion
Warehouse	Radio wave reflection by metal shelves

⑤ Go/No-Go decision: Three options

judgment	conditions	Next action
Go	KPI achieved. Business plan approved.	Moving on to the mass production design phase.
Pivot	Partially failed to meet KPIs	Revise the hypothesis and run the Proof of Concept again.
No-Go	Technically difficult, not cost-effective.	Record the results and withdraw.

The PoC results are compared with KPIs, and the decision is made to proceed as follows: Go (to mass production design), Pivot (re-PoC), or No-Go (withdrawal).

Three common failure patterns in hardware proof-of-concept (PoC).

Here are three recurring failure patterns observed in Proof of Concept (PoC) projects that Macnica MonoCon® has supported.

Failure ①: Attempting to proceed to mass production while still in prototype form.

Raspberry Pi is widely used for proof-of-concept (PoC) and initial prototyping, and there are many cases where companies say, "We were able to create a prototype with Raspberry Pi, so let's mass-produce it."

However, when directly incorporating a commercially available Raspberry Pi into a mass-produced product, it is necessary to individually verify the connector configuration, power supply and heat dissipation, EMC/wireless certification, long-term supply of components, ease of assembly, and maintenance requirements. For this reason, in the mass production stage, it is not uncommon to choose to transition to a module and a dedicated carrier board, or to replace it with a dedicated mass-production board. In one case, the transition from a prototype to a mass-production board incurred additional development costs of several million yen.

It is necessary to select components with mass production in mind from the PoC (Proof of Concept) stage. Postponing verification will lead to significant rework when transitioning to mass production.

Transitioning from Pasberry Pi prototyping to mass production.

The transition from prototyping (Raspberry Pi) to mass production (module + dedicated board), and the items to be individually verified during mass production.

Failure #2: The prototyping partner and the mass production partner are different people.

There are cases where a development partner hired for a proof-of-concept (PoC) is unable to meet the quality requirements for mass production. In one project, it was discovered that the contractor used for the PoC phase would have difficulty meeting the requirements for mass production, necessitating the search for a different partner. Changing contractors requires significant effort to transfer design intent and re-establish quality standards.

Failure #3: Getting stuck at "First, bring me something that works."

Even with a large-scale plan to install tens of thousands of units nationwide, if stakeholders have never seen a working prototype, the project will remain stalled because "they haven't brought anything that actually works yet." The greatest value of a Proof of Concept (PoC) is that it accelerates the process of reaching consensus among stakeholders.

Is the order of the business development schedule that your company has drawn up correct?

Many companies create their own PoC → PoV → PoB (Proof of Concept → Proof of Building) commercialization schedules. However, many companies are commercializing hardware incorporating semiconductors such as sensors for the first time. When they are unsure whether the order and timing are truly correct, it is necessary to consult with experts.

For example, let's say a consumer goods manufacturer creates its own commercialization schedule for a new device. The plan is a 30-month plan consisting of preliminary evaluation → basic evaluation → proof of concept (PoC) → business model realization → sales channel verification → mass production design → trial implementation with a partner → mass production and market launch.

At first glance, the plan seems fine, but those with experience in hardware commercialization can point out several "risky points."

Common characteristics of company-specific schedules	Examples of comments from experts (Macnica MonoCon®)
Proof of Value (PoV) testing at partner companies is conducted after the mass production design.	It's too late. This is a typical pattern where real-world data isn't available in time for mass production specifications, and it's only after the product has been developed that it's found to be "incompatible with the actual situation." Pilot testing should be done on a small scale immediately after the basic evaluation, and should be brought forward.
Each process is arranged in series.	Part of the PoV (Point of Value) and preparation for mass production design can be done in parallel. If it remains in series, the overall length will be 1.5 times longer.
Verification of sales channels and methods follows mass production design.	If you start developing the product before deciding on the sales method, you'll end up having to rework things with the packaging, pricing, and distribution requirements. We'll verify the sales channels ahead of time.
The lead time for obtaining certification is not included in the plan.	Certification requires a timeframe of several months. Unless it's started in parallel with mass production design, market launch will be delayed.
The Go/No-Go criteria at each stage are ambiguous.	Simply moving on to the next stage without a clear plan will lead to failure later. The passing criteria for each stage should be defined beforehand.

Macnica 's value lies in its ability to point out what "should be done in this phase" and "is too early/too late" with a view to mass production. If you bring us your own schedule, we will review its soundness and rearrange it into a realistic "sequence, parallelism, and acceleration."

Comparison of Before (Our company: Serial process with PoV and sales channel verification at the end = approximately 30 months) and After (Expert: Parallel process, brought forward = approximately 22 months)

Frequently Asked Questions (FAQ)

Q1. What is the difference between a hardware Proof of Concept (PoC) and a prototype?

A Proof of Concept (PoC) is an experiment to verify whether something is technically feasible. Prototyping is the process of confirming whether it can be produced as a product. A simple prototype focused on functional verification is sufficient for a PoC, but prototyping requires design precision with mass production in mind.

Q2. Can I request a Proof of Concept (PoC) even if it's just an idea at this stage?

Yes. We can help you from the idea stage, such as "I want to do something like this." Specifications and drawings are not required. We even have cases where clients have consulted us at the stage of "We modified a commercially available XX to create a prototype circuit board, but we lack the know-how to commercialize it" (with an annual target of 5,000 units, starting with 100 units).

Q3. Who evaluates the results of the Proof of Concept (Go/No-Go decision) and how?

The PoC design document is used to mechanically evaluate the degree to which the KPIs (measurement accuracy, communication success rate, battery life, etc.) are achieved. There are three options for the decision: Go (KPI achieved → proceed to mass production design) / Pivot (partially unmet → hypothesis revised and PoC again) / No-Go (technically difficult/not cost-effective → withdraw). At Macnica, we agree with the customer on "which metrics need to be cleared to go" before starting the PoC, so the design minimizes inconsistencies in the decision. Even in the case of a No-Go decision, we deliver a report detailing the technical insights and present the next action plan.

Q4. What if I want something that works immediately without doing a Proof of Concept (PoC)?

In such cases, procurement of off-the-shelf products or small-scale customization is more appropriate than a Proof of Concept (PoC). We actually receive inquiries from clients who say, "We don't need development or a PoC; we just want something that works right now." At Macnica, we listen to the situation and then determine whether a PoC is necessary or if an existing solution will suffice before making a proposal.

Q5. If the PoC is successful, can we proceed directly to mass production?

Yes. We provide consistent support from PoC to mass production design, certification, and mass production manufacturing, all handled by the same team. In one project, we provided end-to-end support for each stage, from the first prototype to the second prototype (30 units), then to the mass production prototype (500 units), and finally to mass production (over 10,000 units per year).

Summary

The commercialization of hardware progresses through three stages: PoC (Does it work?) → PoV (Is it useful/profitable?) → PoB (Is it viable as a business?), gradually increasing the level of verification. Success or failure depends on how the overall roadmap is designed. The first stage, hardware PoC, is carried out in the following five steps.

1. Hypothesis setting– articulate the verification items, KPIs, and judgment criteria in advance.
2. Technology selection– Start the PCB design process after "getting a rough idea" using an evaluation board.
3. Prototype creation– prioritizing "working" over appearance.
4. Field testing– verifying the product outside the development lab, in a real-world usage environment.
5. Go/No-Go Decision– Don't leave things unfinished; clearly define the next action.

It's crucial to design a single business development schedule that goes from having an idea but no working prototype to testing it with a working prototype (PoC), confirming its value (PoV), and finally mass-producing and launching it as a business (PoB). If you have any doubts about the schedule you've drawn up yourself, consulting with an expert is the quickest way to resolve the issue.

Macnica 's PoC development team is comprised of veteran engineers and handles several PoC development projects every year. The team's greatest strength is its neutral component selection ability, which is not tied to any particular manufacturer. While typical device manufacturers promote their own products, Macnica, as a semiconductor trading company, is familiar with products from multiple manufacturers, allowing it to select the "optimal sensor module for the customer's requirements" from a neutral perspective.

Even if you have no experience in IoT device development or DX, rest assured.We will support youfrom proof-of-concept testing to mass production aimed at market launch with our manufacturing consulting service (MonoCon®). Please feel free to contact us if you have any problems.

Site Search