There are 3 major hurdles with prevents everyday robots from being present:
1) Supervising: When it is fully autonomous, things can get very nasty. So the robot should follow the human supervisor.
2) Usablity: Most DIY robots face the problem, that they are made for specific purpose and thus really hard to generalize. Therefore, a thing-manipulation-system will be implemented, whose target can be exchaned modulary.
3) Reproducibility: As far as possible, the parts will be 3D printed, to be able to replace the worn out parts and design new modules without to replace the whole system. And 3D printers are more accessible (also through hub-services) so I can get replacement parts in case of emergency.
With all these aspects in mind, there will be enough critical factors to lay a ground-framework for diy-mobile robots and the people can start implement the "dogginess" of the robot-dog instead of redesigning other narrowly purposed platforms.
Let's face it: Every maker and hacker has dreamed about at least once having a robot companion. But why don't we have them? Because there are some key features missing: - Supervision problem: If a robot is fully autonomous, it can get pretty dangerous. - Usabilty problem: If the robot is not general purpose enough, then it will end up eating dust in your garage, if it is too general, it cannot be used practically because you may be needing in other projects.
- Replacabilty: A custom robot part can be quite expensive or requiring special tools to create or just elaborate. If the parts are 3D printable, you can get replacements easily because nowadays there is enough possiblities for access.