First, the algorithm<->hardware tradeoff. There is often a trade off between the quality of your algorithms and the quality of your hardware. On one hand, you can use an algorithm with simplifying assumptions for speed and performance, but this may require an expensive sensor. However, if your algorithm is robust, it may run more slowly, but you can use a lower-cost sensor. The issue is that this choice has to be made very early in the project with little knowledge of how the design may iterate. If the wrong choice is made, or the technical requirements change, this can result in high cost. The MM-iSAMv2 technology allows you to start with the simplest available hardware while you develop and test and only upgrade when you need. The underlying algorithms leverage factor graph and manifold mathematics, enabling it to solve the most challenging problems without compromising performance. In the end, you can engineer for cost rather than be stuck with whatever you started with.
Second, the true cost of implementation. Once you get everything working in the lab or on your desktop computer, how do you bring that to market? There is a huge gap between a proof-of-concept and a running production product. NavAbility is providing a ready to use cloud platform that can be used from initial concept to scaled-up production system. This way you can focus on your specific product rather than how to host your computations.
Third, the distributed computing problem. The MM-iSAMv2 technology allows for data syncing between edge devices and cloud allowing for the flexibility to run large compute in the cloud and share that data with the edge. This reduces the overall cost of the compute hardware on the edge if your application has any sort of connection to the cloud.