Leading An AI Revolution in Optical Design

Creating Optical Design Scalability with AI

Like most optics companies, we faced a limitation in our company's growth due to finding qualified designers with the right skills and their ability to get the required security clearances. Peak needed a new way to scale our design capabilities and create designs that provide innovation and can be delivered faster with a higher level of manufacturability. As we looked across the commercial and open source landscape, we could not find the right solution for our technical and business requirements. As McKinsey identified in their report, the scarcity of optical design talent in the market led Peak Nano to explore leveraging AI to scale our existing talent. How can we transform the optic design model to drive innovation in our markets? This led us to create HawkAI, which can overcome the design challenges and limitations of commercial optic software market tools, while also ensuring operational compatibility, allowing us to share final designs and perform any required modifications with our customers using the tools they currently employ.

Protecting IP and Security - AI Elephant In The Room

At Peak, we understand the importance of protecting IP and ensuring secure data and access. That's why we deployed HawkAI as an “air-gapped” closed system, ensuring the highest level of security for our most sensitive projects in defense and industry. When we share design files, we use secure channels with our customers who support ITAR, CMMC, and other security standards. We designed our AI system using the same principles for security, system access, and data sharing that are required for advanced defense technology and applied that to our commercial efforts. This commitment to security and compliance with industry standards is a key feature of HawkAI that sets it apart from other optical design platforms.

What's Next for Optic Design?

This blog will lean heavily on the article of Dr. Rich Lepkowicz, “How AI Is Powering the Next Wave of Tech Innovation in Optic Design,” published in the AI Journal. In many ways, the fundamentals of the optical design process have remained essentially unchanged for over 100 years. Even with the advent of tools like CODE V and Zemax OpticStudio, the industry's dominant optical design platforms represent technology essentially unchanged, as the process used for over 100 years. Both tools force designers into the same manual, iterative workflow that has defined optical engineering for decades, searching through reference libraries, patents, and textbooks to piece together solutions based on human memory and experience. The process begins with designers translating broad requirements—field of view, lens count, size, F#, cost—into detailed specifications, then running countless simulations and optimizations manually. Each step requires PhD-level expertise, making the design process slow, labor-intensive, and heavily dependent on a limited pool of highly skilled professionals.

CodeV vs Zemax: Two Sides of the Same Outdated Coin

Industry professionals recognize apparent differences between the legacy platforms. CODE V offers superior optimization algorithms and faster tolerancing capabilities, with better constraint handling for complex parameters like telecentricity and distortion. Its command-line interface and scripting capabilities make complex operations more accessible. Zemax OpticStudio provides a more intuitive, GUI-based interface that many find easier to learn, though this comes at the cost of optimization power. Zemax OpticStudio serves as an entry-level package for less experienced designers, while CodeV targets high-end, demanding applications. Yet both platforms share fundamental limitations: they're built on decades-old architectures, lack modern AI capabilities, and require extensive manual intervention at every stage. As one expert noted, "Complement mature tools like CODE V and OpticStudio with AI-native components to define a modern optical design stack."

What is Required to Move Optical Design into the AI Age?

Moving optical design into the post-AI world requires a new way of thinking about the problem, which involves building a bridge to the past and leveraging the insights gained from previous optical work. Traditional optical design relies on decades-old software architectures built before the GPU era, making it essential to rebuild the entire technological stack with AI-first principles. This means replacing legacy tools like CODE V and Zemax OpticStudio with platforms leveraging modern GPU acceleration, cloud computing capabilities, and distributed processing power that can handle millions of design iterations simultaneously. While we agree with the spirit of Ansys's blog linked above, achieving faster results with GPUs requires more than the same process. 

As Jobs Said, “Think Different” and Ask Better Questions

As Steve Jobs famously said, "Think Different". This is the core of our new approach to optical design. We have shifted our focus from looking backwards at previous designs to an outcome-based model that asks different questions. By leveraging the scale and power of AI, we can now deliver a new model of AI design that is not constrained by the past approach. We have revised the type and strategy of our questions to focus on what the overall optical design must achieve. This new approach allows the AI to test millions of permutations to select the design options with the highest probability of success. By leveraging a wider set of refractive indexes, configurations, and design parameters, we can achieve better design, create innovations, get to the answer faster, and most importantly, we no longer require a PhD-level designer to achieve these results. 

Better Questions, Get Better Answers

One of the most fundamental ways ChatGPT and other LLMs have changed the world is by the simple fact it forces us to ask better questions to get better answers, and it can source a broader range of resources to create those answers, it will bring ideas and thought we have not considered to the fore and challenge assumptions and create new thoughts the lead to better question and results that are improvements over what we would have made in the silos of our "experience bubbles." At Peak, we recognized the need to extend our scope of experience, which led us to develop HawkAI. This platform challenges our designers to ask their design questions in new ways, focusing on the outcomes rather than the traditional design models. We now test our assumptions, objectives, and educations against a broader range of parameters, configurations, and materials, enabling us to “think different” and achieve more accurate answers.

What Are The AI Advantages Over Legacy Tools

The transition from traditional optical design methodologies to AI-powered platforms represents the most significant paradigm shift in the field since the introduction of computer-aided design. While legacy tools like CodeV and Zemax have served the industry for decades, they remain fundamentally constrained by manual, iterative processes that require extensive PhD-level expertise and limit design exploration to human memory and experience. AI-driven platforms like HawkAI are breaking these constraints by automating complex calculations, enabling massive parallel processing, and democratizing access to advanced optical design capabilities. This technological leap is not merely an incremental improvement but a complete reimagining of how optical systems are conceived, optimized, and validated, transforming six-month design cycles into six-week sprints in our internal programs while simultaneously expanding the range of possible solutions beyond what any human designer could manually explore. 

• Massive Iteration Capability - Unlike CodeV and Zemax, which require manual testing of limited design variations, HawkAI can evaluate millions of design permutations simultaneously, testing combinations of materials, refractive indices, and lens configurations at an unprecedented scale.
• Mathematical Optimization - HawkAI leverages Maxwell's equations and advanced mathematical models to optimize designs, rather than relying on human memory, experience, or limited database searches that constrain traditional tools.
• Democratized Access - Traditional platforms require PhD-level expertise, but HawkAI's automation empowers master's and bachelor 's-level engineers to contribute meaningfully to complex optical design projects.

Peak Nano: Pioneering AI-Driven Optics Solutions

Peak Nano's HawkAI technology is a prompt-driven AI design tool that leverages machine learning to test millions of permutations of optics principles, optics lenses, design configurations, and materials, creating optimal lens prescriptions for specific applications. 

• 75% faster design cycles - Reduces traditional 6-month design timelines to 6 weeks, with potential for further reduction to days or hours.
• Real-time optimization - Tests millions of optical design permutations simultaneously using AI automation.
• Instant iteration capability - Eliminates the need for manual searching through reference libraries, patents, and textbooks required by legacy tools.
• Prompt-based design interface - Designers specify desired outcomes through natural language, letting AI generate novel solutions from scratch.

• Superior color resolution - Provides enhanced color clarity and management for sharper, more accurate imaging.
• Wider field of view capabilities - Enables broader FOV designs with better performance than traditional glass optics.
• Complete focus across FOV -Maintains high resolution throughout the entire field of view, unlike some glass optics.

• Lower development cost - Dramatically reduced design timelines translate directly to lower engineering costs.
• Faster time to market  - AI excels at rapid iteration; breakthroughs still require human insight.
• Eliminates manual bottlenecks - Automates complex calculations that traditionally required highly skilled (and costly) personnel.
• Scalable resource allocation - AI serves as a flexible, shared resource across multiple projects simultaneously.

• Massive design space exploration - Evaluates vast ranges of materials, optical architectures, and surface types beyond pragmatic human capability.
• AI GPU architecture - Built on modern computing infrastructure vs. pre-GPU legacy platforms like CodeV and Zemax.
• Secure U.S. supply chain: Manufactured domestically with allied nation sourcing for defense applications.
• Reduced tech debt: Modern platform eliminates legacy infrastructure costs and limitations.

Integrating HawkAI With Traditional Optical Design Tools

Peak Nano's HawkAI platform demonstrates sophisticated interoperability capabilities through Dynamic Link Library (DLL) integration that enables seamless backward compatibility with legacy optical design tools like CodeV and Zemax. The platform can generate design files in formats natively readable by these established software packages, allowing optical engineers to leverage HawkAI's advanced AI optimization while maintaining existing workflows and toolchains. This backward compatibility feature addresses a critical industry need, as many organizations have significant investments in CodeV and Zemax licenses, training, and established design procedures that cannot be immediately abandoned. By providing DLL interfaces that translate HawkAI's AI-generated lens prescriptions into the standard file formats used by legacy platforms, Peak Nano ensures that engineers can benefit from the dramatic 75% reduction in design time while preserving their ability to perform final validation, tolerance analysis, and manufacturing documentation using familiar tools. This hybrid approach represents a strategic bridge between the AI-powered future of optical design and the practical realities of current industry infrastructure, enabling organizations to gradually transition to AI-driven workflows without disrupting ongoing projects or requiring immediate wholesale software replacement.

AI Will Drive the Future of Optical Design

Artificial intelligence represents the inevitable future of optical design, fundamentally changing a field that skilled designers and manual processes have constrained for over a century. The convergence of massive computational power, advanced algorithms, and the industry's critical need for faster innovation cycles has created the perfect conditions for AI to scale and accelerate design processes. While traditional tools like CodeV and Zemax served their purpose in the pre-GPU era, they are now bottlenecks preventing the industry from meeting exploding demand across defense, automotive, healthcare, and emerging technology sectors. By shifting the optics design model to a question-based LLM-type interface, we can increase access, innovation, and time-to-market for optics designs to meet market demands.

Peak Nano is leading the optics design revolution by changing how engineers interact with optical design through HawkAI's prompt-based interface. This allows designers to specify desired outcomes rather than manually searching through reference libraries and legacy solutions. This approach represents a quantum leap from asking "what has been done before?" to "what is possible now?", enabling the exploration of millions of design permutations that would be impossible through traditional methods. Peak's commitment to secure intellectual property protection, controlled data access, and synthetic data generation ensures that companies can leverage AI capabilities while maintaining competitive advantages and regulatory compliance. The platform's ability to test vast combinations of materials, refractive indices, and lens configurations through mathematical optimization rather than human memory represents the democratization of advanced optical design capabilities.