Week 1: (21st June - 27th June)

• Courses resumed this week
• Fixed the crashing problem and set my Raspberry Pi to collect data for about 3 weeks
• Made several code improvements
• Got GPIO output working, got a LED from Ulrich to test it was working
• Continued working on my Mid-Year Semester Report

Week 2: (28th June - 3rd August)

• Made several code improvements
• Continued working on my Mid-Year report
• Started working on my Mid-Year seminar
• Worked on a circular array data structure
• Met with Ulrich and discussed the project

Week 3: (4th August - 10th August)

• Spent the first part of the week preparing for my mid-year seminar
• Gave my mid-year seminar at the city campus (7th)
• Continued work on the mid-year report
• Made slight improvements to the program
• Started running an improved stable version of my program to gather data about Ulrich's house, this data will be sample data when I'm improving the decision-making algorithm

Week 4: (11th August - 17th August)

• Submitted my mid-year report (11th)
• Made some more minor code improvements
• Started to develop the intelligent decision-making algorithm, worked on handling errors when system data is unavailable

Week 5: (18th August - 24th August)

• Met at Vector to discuss my mid-year report, there were a few points that Vector wanted to modify slightly
• Had a brief catch up with Ulrich after the meeting, he gave me advice on what part of the program I should develop next
• Begun modifying my mid-year report
• Continued working on the decision making program
• Implemented a Java class that stores queried system data for each day - known as DayData.java
• Implemented a Java class that holds system data information about many days, this contains many DayData instances

Week 6: (25th August - 31st August)

• Met with Ulrich to discuss my work on the project and my modifications on my mid-year report, Ulrich was pleased with my progress and advised me on what to develop next
• Continued working on the decision making program
• Implemented a Java class that stores rule information, also implemented a new API-using method that produces an instance of this class
• Implemented a Java class that stores condition information about rules, multiple condition instances are stored inside a rule instance

Mid-Semester Break Week 1: (1st September - 7th September)

• Continued to develop the decision making program
• Developed a rule set Java class which holds many different rule sets
• Developed a state machine Java class which will be used to simulate a system, this class needs a lot of work
• Met with Ulrich to discuss project progress
• Ulrich and I found that there is insufficient information to do state machine simulation, we defined more simple decision criteria

Mid-Semester Break Week 2: (8th September - 14th September)

• Improved some existing code - the class that holds recent sets of data
• Implemented new decision criteria
• Set the Pi to run the program again with the new decision criteria
• Discovered some bugs in the program while it was running on the Pi, fixed the bugs and set it to run the program again

Week 7: (15th September - 21st September)

• Found bugs in my implementation of the decision criteria, I fixed these up and let the program run more

Week 8: (22nd September - 28th September)

• Met with Ulrich and looked at the results from the new decision criteria, we found it wasn't working as we should so we modified it
• I implemented the changes and let the program run more, it now also logs more information for debugging
• Daylight savings occurred

Week 9: (29th September - 5th October)

• I discovered my code won't handle daylight savings correctly
• Begun to fix the daylight savings problem
• Begun to develop a php script to update program parameters
• Altered my program to read and process the configuration from the php-generated config files
• Improved the clarity of decision code

Week 10: (6th October - 12th October)

• Tidied my existing parameter loading code
• Changed the decision-logic code to take multiple loads into account, loads decided on based on priority
• Improved the printing/logging of load information that occurs when decisions are made
• Improved the clarity of how the discretionary load information is stored in memory

Week 11: (13th October - 19th October)

• Cleaned up some unused classes
• Overhauled DayData and DataRecord classes to make them more flexible with gaps in data
• Altered the format of CSV logging files
• Further worked on code to adjust log files for daylight savings
• Implemented new GPIO functions for dealing with different loads
• Worked on Javadoc documenting
• Met with Ulrich at his house and tested my program on the device that was set up to switch loads. After a few modifications it worked correctly, however there was a problem with the hardware setup that Ulrich needed to fix
• Worked on the slides for my final seminar
• Worked on the final report

Week 12: (20th October - 26th October)

• Continued to prepare for the final seminar
• Met with Ulrich again and tested my program on the device, it successfully turned on a load
• Gave my final seminar at the city campus (22nd), this included giving a successful live demo of the device in action
• Submitted my slides to Vector for approval, I will upload these slides to this website later
• Worked on the final report

Study/Exam Week 1: (27th October - 2nd November)

• Finished working on the final report
• Successfully submitted the final report to Ulrich and Mano (28th)
• Submitted the final report to Vector for approval, I will upload the report to this website later

Study/Exam Week 2: (3rd November - 9th November)

• Vector approved both the final report and the seminar slides, I have uploaded them to this website
• Project is fully complete