Our Research

ISR and Security: The core responsibility of the Intelligence, Surveillance & Reconnaissance (ISR) community is to deliver actionable intelligence to decisionmakers in a timely manner. To meet current and future ISR demands, Lynntech is developing a diverse suite of artificial intelligence (AI) tools and capabilities designed to complement existing intelligence efforts and incorporate future open-source intelligence (OSINT) capabilities. Areas of expertise include synthetic data generation, object/feature identification, and activity determination. Additionally, Lynntech has extensive knowledge and expertise in fortifying and characterizing AI networks against evasive attack and data poisoning approaches.  Tell me more…

Navigation and Tracking: Modern Infrastructure is reliant on the Global Positioning System (GPS) to provide accurate positioning, navigation, and timing (PNT.) The reliance on GPS signals, which are low-powered and unencrypted, creates a vulnerability in our infrastructure. Lynntech is developing a sensor fusion approach to patch this vulnerability, including a platform that uses inertial measurements units, multispectral cameras and additional sensors to provide PNT capabilities in GPS-challenged environments. Tell me more…

Health and Life Sciences: Development and optimization of neural networks to group, sort and classify raw data; catalyzing solutions well beyond the capability of machine learning models. Numerous Lynntech projects involve the synthesis of deep learning technology with novel computer vision capabilities to answer emerging needs in national defense, counter-defense, public safety and countermeasure development. Tell me more…

Core Technologies: The core technologies behind much of our work span multiple projects and multiple domains. These technologies take the form of techniques, methods, processes, algorithms, computational and simulation tools which we utilize every day to help further our research and commercialization goals. Tell me more…

Our Team

Intelligent Systems is a highly collaborative team of scientists and specialists with experience spanning a variety of disciplines, including aerospace engineering, applied physics, bioscience and medical imaging, computer science, neuroscience, and pharmacology. Though our backgrounds cover a wide breadth of disciplines, we are tied together by our common passion for finding exciting and innovative ways to apply state-of-the-art technology and ideas towards solving some of the biggest challenges faced by industry and society as a whole. Many of the motivating problems that drive our research and development come from the public sector where many large open-ended problems await the right team with the right ideas to come along. We work together to evaluate new potential projects and funding opportunities as well as carry out the scientific and engineering work needed to achieve the objectives of active projects. Many efforts include partnerships and collaborations with academia, granting us access to some of the sharpest minds and cutting-edge ideas out there. We often have several such R&D projects running in parallel and each project is coordinated by one of our Principal Investigators (PI) who are responsible for supervising and delegating workloads between projects, and it is common for scientists, engineers, and analysists to be involved in two to three projects simultaneously.

As part of our ongoing research efforts, we continuously evaluate topics from government-released R&D solicitations and write research proposals in response. When batches of solicitations are released by various government entities, like NASA or the Department of Defense, for example, we come together as a team and decide on which ones we should pursue and who will lead efforts on writing proposals to the selected topics. Upon competitive selection for contract award, such proposals become active research projects. With joint support of in-house engineering, program management, and commercialization support we work to bring new ideas and technologies to market. This internal collaboration facilitates quickly advancing new technologies into the physical/testing environment.

Because Lynntech is a small company, anyone working on a project can have a big impact on where the project goes. New team members start off working on individual tasks and filling in singular needs but over time are given responsibility for developing and maintaining larger pieces of projects and owning systems. We encourage our group members’ growth and provide opportunities for taking on more responsibility, eventually leading projects and deciding the direction of future research. Members also sometimes find themselves slowly branching out or changing disciplines as interests change, industries evolve, and certain projects gain more traction than others. Branching out usually takes the form of joining existing projects or working on proposals to start new ones and however our interests evolve, we always have opportunities to explore the latest and most promising scientific developments as they pertain to market and industry needs.

Our Process

The Project Development Cycle: The Intelligent Systems Group follows Lynntech’s tested and verified research and development cycle. We start off by exploring industry domains for open problems that have a strong need behind them and performing literature reviews that help us understand the state-of-the-art in industry as well as the latest developments in academic circles. We then assemble a collection of possible proposal ideas and evaluate their potential as sources for driving technological development within the group as well as the revenue potential that would come out of a successful project. We brainstorm ideas for possible products and prototypes that solve the open problems we’ve down-selected. The most promising ideas transition towards a proof-of-concept phase, where we begin to work on the minimal effort implementation of a test or experiment which would validate our concept. The initial concept is then refined through iterative testing and experimentation.

Once a technological concept has been validated in terms of its technical feasibility, the next step is the development and testing of a prototype. Prototypes differ from a simple proof of concept in that they not only attempt to demonstrate technical feasibility, but they seek to validate that a technology or methodology can be packaged into a product or service. Development and testing of prototypes is a process which works best when potential customers and vested parties are brought in the loop. Feedback from enthusiastic and invested customers groups is critical for ensuring the technical solutions we’ve come up with satisfy needs of customers and bring value to their lives and their businesses.

The Data Generation and Model Development Cycle: Equally important to the development of new models and algorithms is the creation and refinement of high-quality data sets. Quality data is extremely important in building and evaluating new computational tools, and often data sets are developed in parallel with the algorithms that leverage them in an iterative bootstrapping process – better modeling tools are used to create better data which is then used to create better models.

There are multiple validation steps embedded in this development cycle. Synthetic data can come from generative models, simulations, or from data augmentation applied to real data. Before this synthetic data can be confidently included in a refined data set, it must pass scrutiny. Both quantitative and qualitative comparisons can be made against high-quality real data. Sometimes there are well-established metrics for comparison available in literature, or as part of established best practices for performing quantitative comparison. Other times, a significant effort must be given towards establishing and testing new metrics which themselves must be validated as good proxies for measuring the performance sought in a particular application.

A significant effort is also dedicated to setting up faithful simulators which can reliably produce data spanning the problem space of interest. Sometimes simulators can achieve a desired level of realism and quality, however, lack the tools and generalizability that make their application extensible to the domains of interests. Other times, simulators can produce data which spans the classes, or domains, of interest, but the quality does not meet our needs. This data is piped into data augmentation tools which improve data by either enhancing realism or adding problem relevant artifacts and degradation representative of real-world DAQ systems.

Join Lynntech!

Intelligent Systems at Lynntech applies state-of-the-art computer vision, machine learning, data science, and emerging AI methods to solve existent industry problems. Working closely with clients and collaborators, solutions are promptly iterated in an engineering environment where theory and application must consolidate. We’re always looking for talented scientists, specialists, software developers, and engineers that have a passion for research and development to join our team.

The interview process at Lynntech progresses in the following manner.

Resume: You can let us know that you are interested in working with Intelligent Systems by submitting a resume. Members of the team review the submissions, and if a resume is in good alignment with an available position, we will reach out to you and schedule a phone call.

Phone Interview: The phone interview is a chance for our team to introduce ourselves and Lynntech, the types of projects we are currently working, and provide a general overview of our workflow. We are interested in hearing about your interests, background, work and/or academic experiences. If candidates and Lynntech agree there is a good fit for the position, we forward our hiring assessments with the candidate.

Assessments: Every job candidate is given a non-technical assessment which measures a candidate’s fit with our group culture and their “cognitive agility, behavioral adaptability, and interests”. Once we have received your submissions we will review and discuss scheduling a second interview. This initial assessment is then followed by one or both of the following technical assessments, depending on the position.

• For technical and specialist roles, such as an intern, AI/ML engineer, data science technician or analyst, we ask that you perform a technical take-home assessment. The take-home assignment has a mixture of math, programming, and short answer questions on relevant research and engineering topics.
• For Scientist positions, in addition to the technical take-home assessment, we also request candidates write up a 1-page abstract in response to a past research solicitation (prompt is provided). Proposal writing is one of the core responsibilities of scientists at Lynntech, so strong technical and persuasive writing is highly valued. Scientists at Lynntech frequently communicate ideas and findings to funding agencies as well as to each other and it is important to be able to do so quickly, and in a persuasive manner.

On-site Interview: The on-site interview process allows you to meet our team members and gives us a chance to work with you on a relevant topic.

• For technical roles, depending on the position, we may have you work on a set of “open book” coding problems. Data science candidates might be presented with data wrangling problems. An AI/ML specialist might be asked to import a dataset, and train and evaluate an appropriate AI/ML model.
• For Scientist positions, we ask that candidates present a scientific/technical talk to a general technical audience. We are not just interested in the theory behind your past work – we are asking for a personal narrative. You should be able to tell a story that weaves in how you approached the research/problem, how you worked with others to achieve your successes, challenges you faced along the way, and even times when your hypotheses failed. Depending on your background, we may ask for an appropriate technical demonstration.