The High Cost of Delayed Optical Design: Why "Early Intervention" Matters
In optoelectronic product development, optical systems are occasionally scheduled for the latter half of the development cycle. This reflects an industry reality—optical engineers are far fewer than electronic and software engineers, and project leaders often come from circuit or software backgrounds.
When a team's core competencies are concentrated in one domain, the decision-making balance naturally tilts toward familiar territory. Some believe optical challenges are manageable; others consider circuits or software the core difficulties; still others, lacking optical expertise, choose to postpone it. However, when other systems are finalized without optical constraints, the team later faces a dilemma—either sacrifice core optical performance to accommodate the structure, or start over.
The following two real-world cases reveal the substantial costs of delayed optical design.
1. Case Study One: The AI Project Misjudged by Timeline
1.1 Background
A startup secured significant funding based on its leading AI algorithm team, planning to develop a high-end inspection instrument combining optical systems with AI algorithms. The core team members all had software and circuit backgrounds, and their previous high-difficulty products had been successful. Before project initiation, the team learned that mature optical instruments of this type existed in Europe, and two Chinese companies offered lower-performance versions. They concluded the optical system would not be particularly challenging. Therefore, after project launch, they prioritized core R&D resources for algorithms, planning to integrate optical solutions once the software framework matured. Meanwhile, the company signed a strict betting agreement with investors, requiring national certification of the fully self-developed product within two years.
When the algorithms and other hardware were nearly complete, nearly a year had passed. Only then did they begin seeking optical suppliers. When they approached Haoge Optoelectronics, only 7 months remained before the betting agreement deadline. However, the technical requirements shocked us: they demanded the optical system's resolution and volume directly match flagship products from international top brands (such as Zeiss).
Zeiss, as a century-old giant in the optical field, achieved its flagship product standards through decades of technical accumulation, countless generations of engineers, and iterative optimization with excellent suppliers. The development of such high-end optical systems is inherently a long-term engineering effort measured in "years." Even for an experienced professional optical team, creating a product matching Zeiss flagships from scratch within a few months is an impossible task.
1.2 Outcome
Ultimately, through negotiations with multiple suppliers, the company recognized it must significantly reduce optical technical specifications. Haoge Optoelectronics delivered a version meeting only the certification "passing line" through emergency technical breakthroughs, simplified design, full production department coordination, and mobilization of all supplier resources. This barely fulfilled the betting agreement requirements, allowing the client company to avoid direct breach of contract. However, due to the core optical performance falling short of expectations, the company lost subsequent funding opportunities.
It should be noted that this version was a expedient measure under time pressure, far from an ideal solution. The company's software capabilities were unquestionable, but by ignoring the long-cycle nature of optical R&D, a once-perfect betting agreement became "tasteless chicken ribs"—they missed the optimal window to become an industry leader.
2. Case Study Two: Molds Scrapped Due to Violating Physical Principles
2.1 Background
An environmental instrumentation company had developed multiple chemical-based environmental detection instruments, with strong capabilities in circuits and software. The company decided to develop an optical-based atmospheric particle monitoring instrument, using a rectangular parallel light beam through the monitoring area and measuring scattered light intensity with sensors to calculate atmospheric particle conditions. The team was confident; circuits and software were completed, and even the housing molds were manufactured before seeking optical team involvement. The requirement sounded simple: "Just use an LED and a lens to collimate the light." However, when we saw the reserved space, we were alarmed—absolutely impossible! Their understanding was: "Place a point source at the lens focal point, and parallel light emerges," citing online sources claiming "LEDs are point sources." But all optical engineers know: there is no such thing as an ideal point source, and lenses have aberrations.
A similar case involved another startup, also needing parallel light for sensors, with extremely limited reserved space. When we explained that "LEDs cannot be treated as point sources in many extreme cases," they proposed: "What about laser diodes (LD)?" We pointed out that this sensor would be used in environments with daily human contact. Once lasers are used as light sources, strict safety certifications are required—information they were unaware of. For example, laser classification certification (IEC 60825-1) corresponds to different safety protection requirements for different classes; product safety certifications (CE/FDA) treat laser safety as an independent review item. This means certification costs range from tens of thousands to hundreds of thousands of RMB (approximately 15,000 to 150,000 USD), with a minimum cycle of 3-6 months. For them, this was not merely a technical solution change, but a comprehensive reassessment of product launch timing, costs, and compliance risks.
2.2 Outcome
The environmental instrumentation company's project ultimately could not "make do with the mistake." After multiple rounds of technical argumentation, they finally recognized the severity of the problem. The result: housing molds scrapped, circuits readjusted, project delayed by 4 months, costs increased by approximately 30%. More regrettably, competitors launched similar products, resulting in loss of expected revenue.
The startup's project was relatively "lucky" because they intervened earlier—molds had not yet been opened. After abandoning the laser solution, they could only optimize the LED approach: relax optical performance specifications, add signal processing algorithms to compensate for optical deficiencies via software, and make structural adjustments to gain approximately 20% additional space. Ultimately, the project was delayed by 2 months, with costs increased by approximately 15%.
3. Conclusion: Cross-System Collaboration
The common thread in the above cases is that optical design was scheduled for the latter half of the development cycle. PMs from circuit backgrounds may consider software the challenge; PMs from software backgrounds may view algorithms as the barrier. Optical systems, being "invisible and intangible," are often assumed to be shortcomings that can be addressed later.
However, optical systems are not "gap-filling" supporting actors—they are one of the cores determining product performance. Additionally, optical components have fixed processing cycles. For glass lenses: material selection, blank production, rough grinding, fine grinding, polishing, centering, coating... generally requires 30+ days. For plastic lenses: mold design, mold frame ordering, mold core production, trial molding, inspection, mold revision... also requires 30+ days. These processes are interlinked; each step requires time. If encountering high-difficulty lenses, or inevitably needing special optical components requiring international procurement—such as high-precision gratings, aspheric lenses, polarization beam splitters—the timeline extends further.
Final Recommendations: Whether you are working in AI, environmental protection, healthcare, or consumer electronics, remember these two cases from day one of project initiation. If you are a project leader, next time you hold a project meeting, remember to seat the optical engineer in the front row. If you are a structural engineer, ask one more question before mold opening: "Has optics been confirmed?" If you are a manager, stop treating optics as a "late-stage optimization item"—treat it as an "early-stage mandatory item."
All optical professionals aspire to be your project's early "mine sweepers," not late-stage "firefighters."
Danyang Haoge Optics
丹阳浩格光电科技有限公司