You're striving for innovation in robotics. How can you ensure reliability in your solutions?

I agree, design is an important consideration and the first step is to understand why you are building that particular robot. Take a case scenario where you would really like to have a 360 coverage for SLAM but it seems that a simple monocular camera would do the job. In that case you know what to design for. Hence, building a detailed spec sheet would greatly aid in design. This can then be converted to a rough CAD and explorted to format supported by the sim of your choice e.g. Gazebo, Isaac sim, etc.

I’ll agree with the design for reliability, but that can be subjective. Design engineers often lack the real world, daily use aspect of products. If you design for manufacturability then you (1) gain insights during design & design review to how to make it easier to build a product, and (2) the design is inherently less likely to incur manufacturing deviations from acceptable tolerances. Designing for reliability is not enough…there must be manufacturing for reliability, as well, and that begins during the product design process.

Innovation in robotics must be balanced with reliability to ensure successful solutions. Firstly, thorough testing and validation protocols should be implemented at every stage of development to identify and mitigate potential issues. Employing redundant systems and fail-safe mechanisms can enhance reliability. Additionally, regular maintenance schedules and remote monitoring can preemptively address any emerging issues. Finally, fostering a culture of continuous improvement and feedback loops ensures that reliability remains a priority during the innovation process.

We already have nanometer precision robots in the market. Still, robotics has not reached the popularity of Gen AI. There are still environment factors involved which is the uncertainty in robot design. Improve the design during every iteration. It might take years for a big leap of improvement but the more you improve the more benefits you get in the future.

During my early days of programming, there were many instances where the lack of version control was a huge problem. A small change that I make to my project makes it troublesome to work with. But due to the lack of a previous stable version, the entire project will have to be stopped until I finish the work. Version control systems cannot be underrated, each and every change made to the pushed to git with appropriate comments and a stable tested version of the software must always be present.

I agree, the spec sheet should also be detailed enough to provide idea on what is the environment like in which the robot is going to operate. If you are building a robot which you need to operate remotely, your software needs to be robust since you might not be able to access the location every now and then to debug. During my time in industry, we use to have two system approach. One was in-house for testing our software extensively and once it cleared certain threshold of reliability we would deploy it to the site using CI/CD pipeline. Someone who is either just getting started or have experience, I would strongly recommend containerization of your software.

A lot of realtime capabilities depend on connectivity to a central server. To mitigate this, the reliance will then shift edge computing. Hence, the architecture must support both edge and cloud interoperability.

Software is getting a major update and AI is pushing the need for much more sophisticated software coding in robotics. I had done a project which was done in 2018. The code was cleaner back then, but now it needs a complete revamp. Hence writing a reliable and future-proof coding is needed.

Ensuring software stability is crucial for developing reliable robots, as the software acts as the brain of the system. I prioritize writing clean, well-documented code and following best practices in software engineering. This includes using rigorous version control and conducting thorough testing, from unit tests to system tests. Automating these tests helps quickly identify and fix bugs. Additionally, using real-time operating systems (RTOS) is essential for applications where timing and reliability are critical. These practices ensure that our robots are dependable and resilient in real-world applications. From my project experience, one can also opt for Soft-realtime capabilities as it really helps in achieving high precision tasks.

One of the downsides of electronics is, it need proper maintenance and storage in the proper environment. Even high-priced parts get easily damaged over time. This is one of the main reasons of slow development of robotics

Les composants de qualités ,c'est important d'avoir des composants de qualités mais le problème est que en électronique la plus part des composants sont performant ou fiables dans certaines conditions ce qui pose souvent un grands problème sur le choix des composants mais pour le reste il nécessaire de s'approvisionner chez ceux bénéficiant une réputation ayant des composants de qualité et fiables

Preventive maintenance is the best strategy for the longevity of any product regardless of size. Follow the safety regulations mentioned in any product catalog will a step-by-step guide for preventive maintenance.

In the best case there is no additional training needed for doing the standard task without increasing the risk for additional wear and tear. Just for very special applications a training should be needed.

I would usually make a video guide, as it can easily communicate the do's and don'ts to a non-technical person. Make sure to mark individual parts using a number system in case of hands-on training.

Il y a une nécessité de former les opérateur de ces robots à l'utilisation et à la maintenance de nos robots car cela est non seulement bénéfique à l'exploitant à yang toujours sous la main des gens pouvant faire la maintenance rapide mais aussi car des personnes formé garantie une plus grande longévité à nos produits

Continuous improvement is vital because reliability isn't a one-time achievement but an ongoing journey. By gathering user feedback, analyzing performance data, and staying updated with technological advancements, we can continually refine our robots, address emerging issues, and integrate improvements into future designs. This cycle of feedback and enhancement ensures that our robots remain reliable as they evolve. Ultimately, this approach helps create products that can be used out-of-the-box in any setting, making them more versatile and dependable for users.