Static analysis tools can be used to reduce errors. Just as Software Dev, Test, Bug tracking and Release processes are also good.
A step further is the FDA's MDR program - Medical Device Reporting - https://www.fda.gov/MedicalDevices/Safety/ReportaProblem/def... - where medical device manufacturers are required to report to the FDA when there are problems with their devices. Also, users of medical devices (doctors, hospitals, nurses, patients, etc) can file a report anytime. The FDA then selectively investigates.
Here's a new idea: every single product (software, hardware etc) has a way for anyone to file bugs or design requests or upload videos of scenarios where the design fails. The result would be rapidly accelerating cycles of product feedback loops and with techs like 3D printing, mass customization of devices.
A website that collected feedback like this for endoscopes could be used to make sub product lines for lefty's and button placements for smaller hands. And accelerate product cycle times of feedback to new product in the market from 24 months to 5 months.
Interesting idea to preserve data for very long periods of time. Frank Herbert's books had an idea for a technology that preserved anything (not just data) http://dune.wikia.com/wiki/Nullentropy
There was even a character who had a pill implanted under his skin with cellular samples from many people as a fail safe reference for genetic engineering. Like an implanted library.
Could a useful technology be made out of DNA storage of bitcoin like currency or blockchain IP or contracts? How would it work? Have an implanted pill with artificial DNA inside of it encoding that information? To sign a contract you would have to stick a needle in your arm? Or a thief would need to do the same to steal it?
Or is the idea that information would be encoded in your entire body's DNA so that it would pass down through the generations in case of calamity?
The Tesla shot into space carried a backup of Earth's knowledge in case of calamity: https://medium.com/arch-mission-foundation/arch-mission-foun... The idea is that we may find it in a future where we've partially destroyed ourselves. The Arch may give that future human group a 300 year jump on science, etc.
This won't be easy but cheaper entrants are possible. Take off the shelf components and put them together. Then rely on a human to do the more complex bits.
In the 'Vision' section it requires that the Avatar be able to track motion: "Can you correctly follow an object moving on the floor in front of you?"
The hard way to do this is to develop a vision system with neural nets to track relative motion of any object within the field of view. Objects need to be classified, identified and tracked. Further, the purpose of the object must be divined. That is extremely difficult.
But, if you allow for a wireless connection with a human in a suit and the human operator's brain is doing all of that work (the work a child does when catching a ball), then the challenge gets much much easier. Pass the data to the human (visual) in close enough to realtime so that the human can react while wearing a control suit. Then pass the control suit reactions back to the robot. The problem becomes one of data compression, latency in the system, UI/UX design on the human side of things, etc.
The point is, cheaper, less comprehensive systems could be built that could functionally complete the challenge and win the prize. And those systems could have immediate uses.
For example, what if I have elderly parents who have a robotic assistant at their home fulltime. Some tasks are easy to program (ex. changing a bedpan). But other tasks are much more difficult: a sponge bath to prevent bed sores. The robot could change the bedpan whenever it needed to automatically like a roomba doing the vacuuming. But when it needed to do the bath, it could call me up at home, I could put on my control suit, connect over the internet, and bath my parent from thousands of miles away. The complex 'thinking' and task management of the robot is handled in my brain.
But the basic business case is met: work can be done by a robot with a human 'guide' or 'pilot' at a distance.
The more each marine knows and the more tools each has at his/her disposable, the more flexible each individual and unit will be. Flexibility can be important in fighting.
A Marine unit with multiple materials and multiheaded 3D printers could print more than replacement parts for downed war fighting machinery. Specific weapons or other gear could be adapted to fit the situation.
For example, a particular part isn't working (like the altimeter in the previous comment) due to poor design in bad testing. Marines with access to the raw CAD files of the part could modify the cowling/housing shape to make it more visible. The finished model could be stress tested in a physics modeling program. When it passes, a metal part could be printed and installed. A thermistor could even be added to provide heat to stop frosting. If it works in one plane, it could be duplicated to the whole squadron. And this quick adaptability in changing design method could be applied to anything: backpacks that are the incorrect shape and cause backpain, bump stocks that cause damage on the recoil, or a new water checkvalve screen could be designed to better force water out of soaked boots and stop foot rot. The design work could even happen remotely back on the mainland with more design resources.
This could happen within 24hrs from a supply unit stationed just off the front lines doing most of the work. At the end of a 6month engagement, the vehicles, equipment and gear could all be substantially different than how they started.
This extreme adaptability could end up being a huge advantage.
Its great that this contest is focused on positive applications of technology. This science fiction project from ASU is related: http://hieroglyph.asu.edu/ The goal was science fiction short stories that inspire and set an example for young minds. Much of science fiction is now dystopian. Each story highlights the potential good that science can do and creates a sense of wonder and longing to learn more and do good things with science and technology.
One central theme is that we have all of the technology we need to solve basic problems on the hardware, software side of things. What we are lacking is political will. Some of the stories tackle this topic: what would new tools look like that would help the political system and make it better?
Two that would be great to see for this hardware hackaday: distributed mesh network internet and hardware vaults that hold blockchain contracts of ownership for everything. Another cool one is the haptic feedback shirt. Looking forward to seeing how this contest plays out.
This is hilarious! But very practical. Buttons are a simple user interface to say one thing. And the phone call on the other side of the system is great.
Google's Paper Signals project allow for a more complex input: voice commands. https://papersignals.withgoogle.com/
I can see demand for tons of customized versions of both of these projects. The business model is broken though. Customized hardware or software is expensive. If you could build a cookie-cutter style app creator using emoji-style software blocks (for those with zero programming ability) and stick with off-the-shelf hardware from other creators (Amazon) or even Arduino kits, you could build a business model to support semi-custom products with a mix of hardware and software for many many uses.
This is really interesting. It would be amazing if it could be run from the repository side rather than from a single local machine only. ie Instead of running on my personal dev machine over all of my local repositories looking for changes that I forgot to commit or stash, it could be run from github to all machines that had copies of the repo. Result: a prompt reminder to commit (on the github server machine) of all remote machines work. The use case is that of a project manager looking for uncommitted work from software dev teams.
A further application of this tool would be to create automated backups of every branch of a repo on every development machine. The use case is again from project management: periodic backups for safety or a backup of everything if there is a contract dispute or one consulting team's contract is ending and a new one is starting.
This is a ton of value for all companies that do software to have this tool expanded to cover both of these use cases. You could definitely found a startup to sell this service to any company that develops software.
Interesting article.
If we take it as a given that software programming talent (main requirement for success of fourth industrial revolution - also an assumption) is mainly genetic or some other inherent aptitude that is spread globally, then the number of African software engineers is vastly underrepresented in Silicon Valley or really anywhere.
That's the set of assumptions that drives these guys: https://andela.com/ They allow tech companies to outsource programming work to best of the best of Africa just like India before and Eastern Europe now.
Maybe they, and companies that believe in the same thesis, can add to solutions that overcome the lack of hardware problem. MITs $100 laptop challenge doesn't seem to have done the job.
A step further is the FDA's MDR program - Medical Device Reporting - https://www.fda.gov/MedicalDevices/Safety/ReportaProblem/def... - where medical device manufacturers are required to report to the FDA when there are problems with their devices. Also, users of medical devices (doctors, hospitals, nurses, patients, etc) can file a report anytime. The FDA then selectively investigates.
But sometimes the FDA is too slow, superbugs in Duodenum Scopes is a good example: https://www.reuters.com/article/us-usa-superbug-fda/exclusiv...
An analog to this problem is scientific researchers not publishing failed studies. Or not publishing 100% of the raw data for successful studies for further analysis (finding bias removing selective sampling errors, etc). https://blogs.scientificamerican.com/guest-blog/why-scientis... https://en.wikipedia.org/wiki/Open_science_data
Here's a new idea: every single product (software, hardware etc) has a way for anyone to file bugs or design requests or upload videos of scenarios where the design fails. The result would be rapidly accelerating cycles of product feedback loops and with techs like 3D printing, mass customization of devices.
A concrete example is an endoscope: http://www.genesis.net.au/~ajs/projects/medical_physics/endo... most are built to be held in the left hand as the right hand is reserved for more dextrous work. But what about left handed people? They need a flipped endoscopic design. Also, the button placements are for the size and spacing of men's hands. What about women surgeons? The buttons are placed uncomfortably far away. https://www.amazon.com/Measure-Man-Woman-Factors-Design/dp/0...
A website that collected feedback like this for endoscopes could be used to make sub product lines for lefty's and button placements for smaller hands. And accelerate product cycle times of feedback to new product in the market from 24 months to 5 months.