|
32 | 32 | "items": [ |
33 | 33 | "Traditional wildlife cameras store images locally, requiring physical retrieval", |
34 | 34 | "Cellular models are expensive and unreliable in remote areas", |
35 | | - "Most capture far more useless data than actionable insight", |
36 | | - "DuckWild filters at the edge—sending structured data instead of raw images" |
| 35 | + "Most capture far more useless data than actionable insight" |
37 | 36 | ] |
38 | 37 | }, |
39 | 38 | { |
|
74 | 73 | "title": "Key Technical Decisions", |
75 | 74 | "items": [ |
76 | 75 | "Reimplemented ClusterDuck in Python instead of modifying upstream C++ for maintainability and reuse", |
77 | | - "Dropped hardware AI accelerators due to power cost with little benefit for solar operation", |
| 76 | + "Implemented fault-tolerant control software with watchdog timers and thread health monitoring to automatically detect and recover from failures, with systemd ensuring reliable system bring-up on boot and after power loss", |
| 77 | + "Dropped hardware AI accelerators due to power cost", |
78 | 78 | "Motion-triggered, event-driven inference so AI only runs when needed" |
79 | 79 | ] |
80 | 80 | }, |
|
93 | 93 | "Responsible for motion sensing -> camera capture pipeline", |
94 | 94 | "Implemented a systemd service to manage dependencies, start the networking stack on boot, and automatically restart on failure, enabling reliable system bring-up", |
95 | 95 | "Contributed to ClusterDuck packet handling and Raspberry Pi compatibility", |
96 | | - "Made power and architecture decisions based on real measurements" |
| 96 | + "Made power and architecture decisions based on power measurements" |
97 | 97 | ] |
98 | 98 | }, |
99 | 99 | { |
|
106 | 106 | "Long-range field testing with multiple nodes", |
107 | 107 | "Public demo video + open-source release" |
108 | 108 | ] |
109 | | - }, |
110 | | - { |
111 | | - "type": "text", |
112 | | - "title": "Why This Project Matters", |
113 | | - "body": "DuckWild shows how embedded systems, networking, and AI intersect in the real world. It’s not about maximizing benchmarks—it’s about making technology usable where infrastructure doesn’t exist." |
114 | 109 | } |
115 | 110 | ] |
116 | 111 | }, |
|
0 commit comments