AI Robo System Prototype
Budget: €120 – €250 EUR
Project: AI-Powered Robo-Uber System – Coursework Implementation
Objective
Develop an intelligent taxi dispatch and routing system using AI techniques, including path planning, constraint satisfaction, probabilistic reasoning, and multi-agent coordination.
Key Tasks & Deliverables
1. Path Planning Improvement (20%)
1a: Analyze the baseline system over multiple simulated days. Evaluate metrics like:
Average revenue per taxi and dispatcher
Total joint revenue
Number of active taxis over time
1b: Implement an optimized _planPath function for taxis using a suitable pathfinding algorithm (e.g., A*, Dijkstra). Justify the choice and compare results with the baseline.
1c (Optional): Develop a probabilistic path planner that accounts for traffic and estimates journey time. Analyze impact on fare cancellations.
2. Dispatcher Optimization (10%)
Modify _allocateFare to maximize total daily returns for both taxis and dispatcher.
Consider:
Time to pickup
Number of bids
Fairness in allocation
Compare performance before and after changes.
3. Bidding & Probabilistic Reasoning (20%)
3a: Improve _bidOnFare to maximize taxi ROI. Test in both:
Deterministic (no traffic)
Probabilistic (with traffic) environments
Run at least 3 trials and analyze results.
3b: Write a brief evaluation of the system’s commercial viability and identify areas needing real-world testing.
3c (Optional): Enhance _costFare to minimize fare cancellations using probabilistic reasoning.
Technical Stack & Constraints
Language: Python (presumably, based on skeleton code)
Libraries: Only numpy and pygame allowed. No pre-built AI libraries (e.g., scikit-learn, TensorFlow, GPT-generated code).
Environment: Grid-based world with roads, traffic, intersections, and dynamic fares.
Agents: Taxis and a central dispatcher with bidding and allocation logic.
Deliverables
Fully documented and functional source code with regular Git commits.
A 3000-word report including:
Implementation rationale
Algorithm justifications
Performance analysis (before/after improvements)
Commercial evaluation (for Task 3b)
Declaration of AI use (if any) in the report.
Ideal Candidate Should Have
Strong background in AI search algorithms, path planning, and multi-agent systems.
Experience with probabilistic reasoning and constraint satisfaction.
Ability to write clean, documented code and conduct systematic performance analysis.
Understanding of commercial AI system deployment is a plus.
Objective
Develop an intelligent taxi dispatch and routing system using AI techniques, including path planning, constraint satisfaction, probabilistic reasoning, and multi-agent coordination.
Key Tasks & Deliverables
1. Path Planning Improvement (20%)
1a: Analyze the baseline system over multiple simulated days. Evaluate metrics like:
Average revenue per taxi and dispatcher
Total joint revenue
Number of active taxis over time
1b: Implement an optimized _planPath function for taxis using a suitable pathfinding algorithm (e.g., A*, Dijkstra). Justify the choice and compare results with the baseline.
1c (Optional): Develop a probabilistic path planner that accounts for traffic and estimates journey time. Analyze impact on fare cancellations.
2. Dispatcher Optimization (10%)
Modify _allocateFare to maximize total daily returns for both taxis and dispatcher.
Consider:
Time to pickup
Number of bids
Fairness in allocation
Compare performance before and after changes.
3. Bidding & Probabilistic Reasoning (20%)
3a: Improve _bidOnFare to maximize taxi ROI. Test in both:
Deterministic (no traffic)
Probabilistic (with traffic) environments
Run at least 3 trials and analyze results.
3b: Write a brief evaluation of the system’s commercial viability and identify areas needing real-world testing.
3c (Optional): Enhance _costFare to minimize fare cancellations using probabilistic reasoning.
Technical Stack & Constraints
Language: Python (presumably, based on skeleton code)
Libraries: Only numpy and pygame allowed. No pre-built AI libraries (e.g., scikit-learn, TensorFlow, GPT-generated code).
Environment: Grid-based world with roads, traffic, intersections, and dynamic fares.
Agents: Taxis and a central dispatcher with bidding and allocation logic.
Deliverables
Fully documented and functional source code with regular Git commits.
A 3000-word report including:
Implementation rationale
Algorithm justifications
Performance analysis (before/after improvements)
Commercial evaluation (for Task 3b)
Declaration of AI use (if any) in the report.
Ideal Candidate Should Have
Strong background in AI search algorithms, path planning, and multi-agent systems.
Experience with probabilistic reasoning and constraint satisfaction.
Ability to write clean, documented code and conduct systematic performance analysis.
Understanding of commercial AI system deployment is a plus.