New to the forum.
I'm not a developer, but a 'packager' of technologies and tech concepts to explain and demonstrate what is possible. Usually for manager-types.
I'm looking for help understanding if my idea is feasible, and if yes, then possibly some (lightly funded) help making it real as a prototype.
Idea flow:
Set-up:
- Analog clock on wall about 10 feet up (time is... 3:46). It's on a shelf so you cannot see time from ground.
- Tello on the ground. in front of the clock, but camera facing away.
Action:
- Tello launches
- Either:
+ Goes up to pre-programmed height of 10ft
or
+ executes a corkscrew scanning manuever where it is spinning and rising "looking for clocks"
- At 10 ft, spins/scans for clock faces
- Locks on to clock face
+ if we get fancy it could move forward a little, or 'wiggle' to indicate it has locked on
- Image processing reads the clock face (perhaps using: Deep time (using Tensorflow to read clocks) · Felix Duvallet or some other ML/DL pre-train)
- Outputs the time e.g. 3:46
- Lands back at home
The purpose of the demo is to show that Ladder climbs can be avoided for certain inspection and data gathering tasks using 'low cost, off the shelf' stuff available now.
The ML part is simply to push the edge and allow a demo of integrating IOT (Tello) with 'mass' ML/DL/AI (Tensorflow, AWS, etc)
Thoughts?
And thank you in advance.
Lomotpk
I'm not a developer, but a 'packager' of technologies and tech concepts to explain and demonstrate what is possible. Usually for manager-types.
I'm looking for help understanding if my idea is feasible, and if yes, then possibly some (lightly funded) help making it real as a prototype.
Idea flow:
Set-up:
- Analog clock on wall about 10 feet up (time is... 3:46). It's on a shelf so you cannot see time from ground.
- Tello on the ground. in front of the clock, but camera facing away.
Action:
- Tello launches
- Either:
+ Goes up to pre-programmed height of 10ft
or
+ executes a corkscrew scanning manuever where it is spinning and rising "looking for clocks"
- At 10 ft, spins/scans for clock faces
- Locks on to clock face
+ if we get fancy it could move forward a little, or 'wiggle' to indicate it has locked on
- Image processing reads the clock face (perhaps using: Deep time (using Tensorflow to read clocks) · Felix Duvallet or some other ML/DL pre-train)
- Outputs the time e.g. 3:46
- Lands back at home
The purpose of the demo is to show that Ladder climbs can be avoided for certain inspection and data gathering tasks using 'low cost, off the shelf' stuff available now.
The ML part is simply to push the edge and allow a demo of integrating IOT (Tello) with 'mass' ML/DL/AI (Tensorflow, AWS, etc)
Thoughts?
And thank you in advance.
Lomotpk