First Steps On a Project Flashcards
First thing you should do when being added to a project that’s been going on for a long time already
Build yourself a “glossary/ontology” and a “Stakeholder contact list”
-you want to be able to understand is being said when important terminology is being used
-you want to know who was involved on this project (could be dozens of people)
-some of them are going to be resources
-each of them will have different expertise and background
-each of them will have a unique ROLE on the project
-some of them have much more authority than you do and you will not be able to share information with them directly
- There will be different protocols for communicating with each of the SMEs (some of them can only be reached under specific circumstances, or only contacted by certain people)
Importance of FIRM deadlines
so that you know when it’s time to move on and actually start working on the next phase (people are MUCH less productive without deadlines -remenber Dr. Lineberry’s 2 stage SLI team)
Getting Mentors
find people who have done this before and get suggestions throughout the design AND the build)
-get them involved early on, show them your model before you get to any real design/build work
-waiting to do this will result in bias: You will think that their feedback is less important/useful than it really is because of the amount of work you will have to do to implement it
-Need to discuss all systems aspects of your design (both inside and outside the SOI: risks, corporate partners, agencies/organizations, weather/environment, etc.)
-ask them about RISKS and failure modes (for each subsystem: what could go wrong and cause problems?)
-you need to plan how to prevent/mitigate each of these risks early on
-if there are enough of these you may need to document them in a FMEA
-some of these risks may require analysis/simulation so that you can quantify the probability and impact (this is especially important for the failure modes of structures)
-make sure they actually get to see the bill of materials, and can confirm that you’re not making any mistakes (always make SURE you are buying the right thing for your application BEFORE you pay for it)
-have them evaluate your “measures of effectiveness” (MOEs) BEFORE you do your tradestudies. There may be something important you should be weighing into your decision.
-always have the mentor double check your work before you push it into production (they may find your mistakes)
6 parts of understanding a software tool that you will be using to help with your design
- Basic Tutorial/Features
- Trouble shooting
- Settings
- Shortcut keys
- Scripting opportunities
- Plug-ins/APIs
Importance of understanding what structural material you have available
-just knowing the electrical components is not enough, you need to know what the structures will be made of so that you can plan out procurement/manufacturing/assembly
-Discuss it with the manufacturer (they will likely have material recommendations or even requirements)
Importance of having pseudocode EARLY
sometimes need to spell it out for the software guys (remember clear, concise, consistent communication)
-Quality matters, make sure the pseudocode is readable/accessible for the software guys (actually ask them to verify that it is readable/accessible and address their questions)
-use comments where necessary
-Use UML to create diagrams to illustrate the structure/behavior of the software
-this the definitely something you would want your mentors to look over
Chosing and getting familiar with your collaborative platforms
-Chosen based on the requirements of the project.
-can literally build a trade study to evaluate this
-Primary decision criteria: efficiency and accessibility (prioritizing “schedule” over “cost” or “performance”)
-only use high-performance feature-heavy software if you need it
-Does it have real-time editing, or not?
-are you going to have to share a floating licence?
-make sure you are asking the SMEs for recommendations
-Just like the simulation tools, you need Tutorial + Trouble shooting + Shortcut keys + Settings
Things to know about your sensor early on
GitHub libraries, sensor fusion, code for reading/writing/parsing data, calibration, application circuits, project videos, breadboarding)
-Need to ask your mentors if they’ve ever used it before (make sure it’s not going to be a suprise pain in the ass)
Importance of understanding the entire manufacturing process early
know exactly how it will be machined/printed/assembled)
-be honest with yourself: If you think a particular assembly/integration/maintenence task might be a pain in the ass to do, there is almost certainly a better design alternative that will make it easier (if you don’t see an opportunity to do that, ask a mentor)
-same applies to disassembly. If you built it, chances are excellent you will have to dismantle it in the future (either because you have to repair it or because you decided to upgrade it). Want to make that as easy for you as possible
- ideally you would get some practice assembling it and disassembling it so that you know how it’s done and that way you do not forget how to do it when it comes time for you to fix it
-make a step-by-step assembly plan called a “Master Integration Plan” (MIP)
-need to build verification testing into this plan: testing performance of individual components (component testing) and testing the functionality of each subsystem including PCBs and welded structures (integration testing)
-sometimes you may need to rent/buy/build some piece of equipment in order to do the integration testing properly (always ask a mentor to make sure)
-if you are not doing the manufacturing yourself, that does not excuse you from doing this step (because you need to make SURE it is manufacturable the first time, otherwise your manufacturer might send back your order and tell you that you need to redesign)
-need to talk to the manufacturer who will be handling it early in preliminary design (ask lots of questions)
-if possible, ask if they can take a look at your BOM
-Make sure you actually have enough space for all your components (without collisions/clipping), plus the additional keep-out space needed to do the assembly, soldering, and maintenance tasks (CAD is the best tool for this)
Importance of good swap-C budgets
-it’s okay to initially BS swap-C budgets (but you have to refine them as you get more detail/clarity)
-Make sure each of these budgets includes a decent margin for “creep” (you will almost certainly consume more volume, weight, power, and money than you originally intended)
-Important to carefully cross-refernce your power budeget v. your schematics/BOM and resolve any inconsistencies between your power budget versus your schematic/BOM (seek help from your SMEs if need be)
- does your schematic actually supply the correct amount of wattage/current/votage to the right components?
-you can use various strategies for proving your design is within your SWaP-C budgets (not all of these will be possible or applicable to any given project):
-Size: can 3D print the parts/PCBs you want and actually construct a mock up. This helps identify potential problems with both your parts selection AND integration
-Power: one of the advantages of through-hole components is that you can breadboard them before designing the board and actually measure things like the current flowing from the batteries, and refine your design accordingly
-if something does not come with a datasheet (like a salvaged motor) or there is no datasheet (like nichrome), that means you’ll have to test it and characterize it’s behavior
-this needs to happen BEFORE you start looking at power supply stuff (otherwise you risk dropping a bunch of money on multiple batteries when you only really needed one, or having to pay for additional batteries)
-if one component requires particularly high power, consider looking into alternatives.
-It’s often easier/cheaper to go keep the power consumption of the system to a minimum than to have to design a power supply around getting enough wattage to one or two power hungry components
Importance of doing ALL necessary trade studies
-they are not just for selecting parts: they can help when choosing how you will satisfy a requirement, meet a use case, or solve a problem
-every time you have a configuration/structure/etc in mind for this project/application, you should look at what other options are out there and compare
-if you skip out on a trade study, you increase your risk of having to backtrack and explore alternatives later on when your assumptions fail (which wastes time)
-and if you still do end up having to backtrack and explore alternatives, at least then you’ll know exactly which one you want to try next and why
- Your mentors will need to see your tradestudies so that they can make sure you’re not missing any MOEs
Four steps of Pre-project research
Step 1. Very preliminary research
-figure out who you need to talk to and what you need to ask them
-get obsessive
-learn everything about the application
-find a good simulation tool and learn how to use it (Basic Tutorial/Features, Trouble shooting, Shortcut keys, Scripting opportunities, Plug-ins/APIs)
Step 2. Interview people who have done (or at least attempted) something similar
-there are NO new ideas. If you look hard enough you will find someone who has done (or at least attempted) this before. Don’t just poke around on Google, actually ask people for introductions/direction.
-you want MORE THAN ONE person, from more that one background: ideally two people who disagree with one another
-specifically ask them for recommendations on resources you can use and research you should be doing
-ask lots of why questions: make them explain to you why what they’re suggesting is the best way to do this
-set up a follow-up date that is realistic and still gives you time to follow through on their recommendations: use it as a forcing function
Step 3. Power through as much additional research as you can
-verify everything they said
-look through all the resources they provided
-don’t just accumulate resources
-do some chatGPT queries to see if there is an even better approach than the one they recommended
-research BOTH: you need to be knowledgeable about the stuff that they recommended you look into but you also need to determine whether or not it’s right for your application
Step 4. Follow up with those people you interviewed after you’ve done all this research
-set a date with them in advance and use it as a “forcing function”
-get validated on your progress so far
-once you’ve really done your research, see if you can get it all published (it boosts credibility)
-validation: ask someone to take a look at it when you’re done so that they can fill in any blanks and correct any misconceptions
Do not wait to start the research on the parts of the project you do not understand
-do it early, do it fast
-Ai tools like chatGPT can help with that (everything it says needs to be verified by mentors/SMEs)
-you need to reach out to people who really know what they’re talking about
-admit when you need more education, Don’t be afraid to take the classes necessary to really understand the stuff
-you need to set a due date, so that you know when it’s time to move on and actually start working on it (people are MUCH less productive without deadlines -remenber Dr. Lineberry’s 2 stage SLI team)
-if you will not be doing 100% of the manufacturing yourself, then this step should also include researching the manufacturer and obtaining a list of their capabilities
There’s no sense in doing the simulation without an accurate understanding of…
THE THEORY!
(Otherwise you won’t know what this simulation software is trying to tell you)
Three ingredients that actually make a project work
- Deadlines
- Accountability
- Lots of help (Delegate + get SMEs)-Try to connect your project to some sort of forcing function: some competition/presentation/event were you intend to showcase your work. It forces you to actually get the work done, like having a due date for an assignment (this is why Terminus, CanSat, and Astra were more successful than HILT or Tartarus)
-get your own mentors and SMES (literally build a network leading up to the project) and
Four reasons for having a partner (not a mentor/SME) to work with
1) Can learn from their experience
2) Promotes accountability
3) Reduced Adjorning
4) Delegating
Always try to do your project with a partner (keep the solo projects to a minimum). Your partner should be someone more experienced/skilled, but not necessarily more knowledgeable (like Josh). Remember, you have the SMEs for the knowledge gaps.
That’s how you actually learn. If you do not, you will miss out on opportunities to learn (especially when it comes to troubleshooting and refurbishing your design)
-doing it with a partner also improves accountability and increases the probability of completing the project (they won’t let you slack off)
Also helps mitigate the effects of the “adjoining” phase of a project. You need to reward yourself for making progress so that you don’t lose interest, energy, or focus
Project onboarding materials
Always read through the onboarding materials when you join a project. If you’re on a project and they don’t have onboarding materials, make some.
Include a glossary or acronym guide, a contact list of contributors or stakeholders as well as their role and the protocols for contacting them,
Applying the critical path method (CPM)
Involves identifying specific tasks that will directly impact the amount of time it takes to complete the project. In general, you want to get these tasks out of the way first so that you can have a better idea of your schedule.
Applying the critical path method (CPM)
Involves identifying specific tasks that will directly impact the amount of time it takes to complete the project. In general, you want to get these tasks out of the way first so that you can have a better idea of your schedule.
Preliminary risk analysis
It’s important to understand the biggest risks to your project performance and schedule. If you are new to a project this is one of the first things that you need to read and get yourself appointed with. If you are on a project and it isn’t there this is something you can volunteer to do.
Importance of perseverance when looking for mentors and SMEs
There are no new ideas: if you ask around long enough, you WILL find someone who is attempting or has attempted the same thing you are trying to accomplish. Keep looking until you find them (more than just Google, actually ask people to point you in the right direction)
-specifically look for people with different backgrounds/perspectives so that one of them doesn’t lead you down the wrong path or leave out important info (SMEs can disagree)
Properly delegating
On projects with people, you want to make sure you actually delegate (which is hard, especially when you want all the experience for yourself)
- Best way to do this is with a tasking list: make a highly granular list of things that need to be done and give your team members an opportunity to volunteer for tasks.
Importance of getting inspiration and its role in planning a project
For some projects you need requirements, use cases, AND inspiration. Look at what others have built to help decide what you like and what you don’t like.
Example: when John was building his workstation, you tried to get him to think about his needs and how the design should reflect those needs. Later you found a design online that never occurred to you and completely blew your mind.
How to avoid bias when researching for your project
Unless you’ve had some of it validated, question everything you know. All your existing knowledge and assumptions fall into three categories:
- True, but incomplete (needs context or clarification)
- True and valid
- Untrue
*This also applies to assumptions and even excuses. Need to carefully evaluate what is in your mind with this framework
The very FIRST step when doing research for your project
Start project research by looking for services specifically for the research you’re doing. If you just rely on generic means of gathering information, you might miss out on actual services or platforms designed to help provide that exact information to people just like you (example: when you started talking with Jon about building your own PC and he directed you to pcbuilder.net after you had spent hours relying on chat GPT)
The fastest way to do research on your upcoming project
The holy trinity of gaining deep understanding of something FAST is:
1. Lengthy video content
2. Flashcards
3. Mentor/Interviews
4. Practice
The first pass on writing the requirements
Need to reflect the goals of the system. Research/analyze the goals, and then convert the results into requirements (almost like you’re writing requirements for a student competition).
