II2302: Sensor Based Systems for 2022

Welcome to II2302 for academic year 2021-2022! This course is an introduction to sensor enabled devices, applications, and systems. The topics that will be covered are technologies, algorithms and design factors that will be used in next generation personal communications, entertainment, security, sensor bots and other emerging fields such as found in the Internet of Things and Machine Learning. Four broad technical areas are addressed:

Sensing and sensor technologies, and their impact on embedded design.
Role of context measurement in embedded systems and applications.
Multi-sensor and aggregation device models and methods.
Design methods for devices that use sensors in order to get the performance you want.

After completing this course, you will take away a knowledge of basic sensor technologies, and their role in information systems, especially future ones that exploit communication. Types of sensors, how they are networked and communicate, methods of context measurement and use, and how the use of sensors will change the direction of future personal and embedded system design are among the material covered.

As the class progresses, this web page will be updated with material from the lectures, pointers and references to selected material, and lab information. Be sure to come out and check the page from time to time, as last minute information, schedule changes, homework assignments and lab information will all appear here!

Schedule:

Course lecture days and project days can change. Please check Schema for the class starting times and locations. If there are any special changes or exceptions to the schedule shown in TimeEdit or Daisy then they are noted below, so be sure to check this web page from time to time to keep current of any schedule changes that might occur.

Discussion and Project Material

Pointers to the lecture slides and other material can be found on this web page. It will be more or less continuously updated, so look for new stuff.

For those of you who want to make your own circuit boards in the lab milling machine for your projects, the design guidelines to do that can be found here.

Projects

The projects areintended to let you work with some of the ideas presented in the lectures, and get you thinking about the sorts of things you would introduce to the world that exploit sensors, devices, services and systems. The best way to look at it is a way to get ideas for future study, work or research ideas and to get hands-on experience in designing and building systems and applications that use sensors in some way. I've been involved teaching similar lab project courses where students have gone on to start small companies based on their ideas. Don't worry too much if you don't recognize all the technologies in the list of hardware, software and development tools that can be part of the projects. You'll learn about them in the Mentorspace.

Sensor Based Systems is an opportunity for you to not only discover the world of sensors, but to also develop and practice applied engineering methods that build on the theory in science, technology engineering and math that you have learned in other courses. Making the bridge between theory and applied practice is very important if you want to go on to a professional career in technology when you leave KTH. The Mentorspace can help you develop those skills. For a very short example list of the sort of things that the world of technical companies think are important, here is some information to think about.

Week 3 Discussion: Course Goals and Context Measurement

In the first part of this week we will talk about the logistics and goals of the course and discuss some examples of why sensors and actuators are needed in future IT systems and products. The take away from the first week with respect to technology is an initial idea of:

How embedded devices and product ideas around them start to enable systems created from such devices. Such systems will be embedded in that their functionality will be vertically integrated around a specific task or task model, for example a sensor enabled ID card.
Technologies to form such systems are very diverse. There are multiple issues surrounding how to perform this integration of technology. These are issues we will study.
Sensors have a very direct role in these embedded systems. In fact, devices we carry daily have electronics in them that either can act as, or support the use of sensors.
The idea behind how the integration and use of devices, services and programs have formed new application spaces that could not be accomplished in any other way. Also, how these sensor based systems are driving further development of the Internet, for example with Machine-to-Machine architectures that form the Internet of Things. In addition to the baseline hardware and software, to do this you need:

Sensors: Both for controling devices and to determine other situational data (context) that will affect how an application will work.
Communication: There is no more single “box”.
Protocols: Standard ways of discovering, describing and communicating.

The impact of Performance vs. Power vs. Money. This will pervade all aspects of embedded system technology. The take away here is that performance, power and money, when taken separately for any aspect of a system, may lead a designer to make poor choices. When taken as a function of each other, a designer gets a better idea of the suitability and value of technology as part of the intended system.

The first part of this week also introduces the role of context measurement in existing and future systems. The role of sensors in such systems is to measure physical phenomena that can then be used to make some inference regarding context that the system and application can use. By the end of the week we will be discussing the following points:

What is meant by being Context Aware.
How context has been measured.
What to do with the data. What’s new that you can only do by measuring context.
The hard, future problems.

There are some assigned readings for this week. They will be used in class discussion and possible future homework. The are:

Understanding and Using Context, Anind K. Dey
Ask Not for Whom the Cell Phone Tolls: Some Problems with the Notion of Context Aware Computing, Thomas Erickson

The first reading is important for understanding the entire idea of measuring context, which sensors make possible. The second reading is fun. It starts to show what can go wrong if one takes the idea of context a bit too far.

Week 3 and 4 Project:

This week we will start on the lab projects for the course, and there is something to hand in at the end of week 4. The following will give you instructions for the project part of the course, and explain what needs to be handed in for both next week and at the end of the project. The idea of the project is to design and implement a system and application that uses sensors, and then to implement the system using mentorspace resources. You have a lot of flexibility in deciding what to do for your sensor based system. Here are guidelines for the project, and what you will need to hand in as a final project report next March. Also, there will be presentations and a demo of what your team has done.

The application needs to use sensors in some way. Context aware applications, or some kind of measurement process are good areas to think about. The best sort of ideas are ones that you think would make a good product, and satisfy some kind of customer want. In other words, the project application should be useful.
What is meant by "use sensors in some way" is to show how you can use one or more sensors to solve an application problem according to specifications and requirements you come up with. You will need to do some research and characterize your problem. Your solution should start with a sensor and show how your project group used it to solve the problem and how well the sensor solves the problem. You will want to collect data from your solution and analyze your results. There are lots of pre-existing sensor solutions that one can just buy, such as solutions from Arduino. However, just buying or cutting-and-pasting a finished solution from the internet and turning it on is not what is meant here by using sensors. Here the intent is to develop your ability and skills to use sensors by developing and characterizing a solution yourselves.
It is OK if you want to use a pre-built microcontroller board for your project like an Arduino. But you should design the hardware that uses the sensors yourself. In other words, don't just buy it or cut-and-paste.
If your application needs some special sensors, let me know as we may have the sensors you need. Right now, just think about interesting applications. Don't worry too much about if we can get the sensors yet and remember that often you can make them yourself.
Applications that use some form of wireless communication are also very good to consider.
You will work in project groups of from 2 to 4 people. Working in a group is ideal as you all have complimentary skills that you can share. Note that you must work in a group. Doing the project for the course by yourself is not an option because of the shortness of time, and because a person working alone can not leverage from the insight and skills of others working closely on the same goals.
The assigned project presentation day for this course is on Wednesday, March 16, 2022. The presentations will be in the Mentorspace and at that time you will also need to demonstrate your project system. You also will need to hand in a full report about your project. The guides for the demo, presentation, and written reports are at the end of this web page.

     After forming your team, the goal will be to define and start working on your sensor based project. In week 4, your group needs to hand in a short proposal that describes your project idea. A typical proposal will be about 2 pages long and in this report you should explain what your application idea is, what kind of sensors it will use, and what problem it will solve. You should also include a high level block diagram or sketch of your initial application solution idea. You don't need a full schematic, code or other details yet. Just diagram the functional parts to help make your idea clear. The purpose of this is to help insure that your project satisfies the goals of the course, and is something that can be done in the time we have. Please email this to me by the end of the day on Wednesday, January 26. That will give me a chance to better understand your project idea before our next class session. Also please be sure to include the names of all the members in the group and email addresses. You hand this in as a group, not individually.

   The presentation on Monday, January 24 is about your project idea. The point of this presentation is to show what you would would like to achieve using sensors, and to be sure that the project your group has decided to do is possible to do in the time we have. The presentation should be take about 5 minutes followed by questions and comments by everyone in the room. If you want to make some presentation slides, then you should have about 3 slides at most. Be sure to include information about what the project is about, how it uses sensors, and how it is useful.

     Starting in week 5 your group will need to do a short presentation every week during scheduled project time as a technical progress report, and as a way to put problems and issues that come up to the entire class so that we can help solve them. You also will need to email to me your weekly technical progress report. It should only be 1 page long, and it should be in the format of a real technical progress report that an engineer in a company typically writes. A good example of a written weekly progress report is shown here. Be sure to ask me any questions you have about this because if you get good at writing tech reports like this, it will really pay off in your future career!

Week 5 Discussion: Signal Characteristics and Conditioning

We start now to look at sensors, signals, noise and some of the conditioning and processing we can do. In a large sensor system, or one that is formed from the aggregation of a lot of devices having sensors, it's likely that the set of sensors being used in an application will be heterogeneous. What taxonomies can we apply here? We also would like to develop a general, system model for noise that can help us characterize a measurement system, and decide if it will perform the way we expect. As usual, these technical considerations will be biased by cost factors and functionality, or lack of it, from the rest of the system. The topics presented here will give us tools to use when we address sensor and application technologies next.

What characterizes the sensor signal types
Accuracy and Precision with respect to these signals
General noise model with respect to these signals
The impact of the processing model

Week 5 Project:

By this week your project group should know what your project for the course will be. You should know what needs to be sensed, and what problem the sensor based system will solve. This week you will want to work out what the performance of the sensor system should be in terms of accuracy, precision, resolution, speed, and system noise. In other words, you should have a complete set of specifications with metrics and units for how you expect your system to work. At this point these specifications are initial, target specifications. Your actual project may or may not reach these goals. That's OK and note that at this point you have not selected actual sensors but rather are thinking about the characteristics that your system will need with respect to sensors. For now, use the information from the discussion sections and come up with estimates for how you expect your system to perform. Be sure to show that data when you hand in your project update report next Friday. Email your project update report to me by the end of the day on Friday, February 4. Name it something like week_5_update_project_name. You can email it to me using PDF, DOCX, or ODT (libreoffice) formats.

Week 6 Discussion: Continuation of Signal Characteristics and Conditioning and Time Varying Noise

Last week we started to look at sensors, signals, noise and some of the conditioning and processing we can do. We will continue with this and develop a general noise model that we can use with sensor based systems. This includes static or very slowly changing noise sources and their affect on accuracy of measurement. We will then extend this by taking into account noise sources that are part of the observed signal, and vary with time. Here we look at the following:

Systematic and random noise. How they are characterized.
How random noise gives rise to less than perfect precision.
How to use the PDF and PE from your data measurements to decide if your measurement precision is good enough.
How simple filtering such as averaging will increase your precision and lower your probability of error (PE).

Assigned readings are:

Statistics, Probability and Noise, S. Smith

Week 6 Project:

Last week your project group determined the performance that the sensor based system you will design should have. At this point you should have specifications for your system with respect to accuracy, precision, resolution and dynamic range. These are target expectations. The next step is to come up with a design that can include real sensors and that as best as you can determine will meet the target expectations for system performance that you decided on last week. In this week's technical progress report you should include the following.

If your system uses commercially available sensors, then you should select a set of sensors that you can obtain that will meet the target expectations that you decided on last week.
If your project is to build your own sensors, then you should come up with an architecture for the sensor. In other words you should decide how the sensor will work and how you are going to build it.
In all cases you should also show a system block diagram of the rest of the sensor based system that you will build. In other words, you need to show the entire system at the functional level for sensors and other hardware that it will use.

Note that you don't need to have your design completely done at this point. You don't need to show schematics or lines of code yet. But by the end of the day on Friday, February 11, you should hand in a technical progress report that includes all the information mentioned above, and explains how you expect the system to work (the theory of operation) and from that why you think what you are designing will meet your target expectations that you did last week. We will meet in the Mentorspace during our scheduled time on Friday and your group should give a 5 minute progress report that explains what you have put into your technical progress report. Remember that this is a good time to ask questions to the entire class to get ideas, advice, pointers, and to share any and all useful information. Email your project update report to me by the end of the day on Friday. Name it something like week_6_update_project_name. You can email it to me using PDF, DOCX, or ODT (libreoffice) formats. Remember that information about how to write technical progress reports can be found here.

Week 7 Discussion: ID Management, Biometrics, and Biosignals

This week we will talk about ID management and start to look at systems where sensors are used to measure biosignals. As we look at methods and systems for ID management, we will include identity tokens, and advanced tokens and systems that will exploit biometrics. From a system level, we will look at how identity is managed and used across a wide spectrum of commercial applications. We will also look at several related issues including issues of legality and privacy. One of the ultimate uses of sensors in ID management is to perform biometric identification of people. Biometrics allows systems to determine who a person is based on physical characteristics intrinsic to the person. This is an area of great interest in the commercial space because practical solutions for many applications are just now emerging. Also biometrics is part of a greater area of sensor use that is concerned with bio-signals. In these applications, sensors are used to measure physiological phenomenon for applications in health care, personal and team athletics, and for important consumer applications such as fitness and well-being. Bio-signals are challenging to measure and involve a whole new world of measurement ideas. If you are interested in bio-signals and related bioinstrumentation related phenomenon, then you will find this week particularly useful.

Week 7 Project:

This week you should have a design for your sensor based system that is based on the overall design from last week and the expected performance that you came up with two weeks ago. Your design should be such that it can be built. If you are designing hardware you should have schematics drawn and ready to use to make your hardware design. For example if you want to make a circuit board, then you should have your design drawn in some circuit board CAD tool like KiCAD. If you want to wire your design onto perf-board, then the schematic can be hand drawn, but it should be complete. If you are designing software or firmware, then it should be completely diagrammed out and you should have started to write the code. If you are planning to make a printed circuit board, you can find the design rules for making PC boards that we can make in the Mentorspace here.

Note that you don't need to have your design completely built and running this point. But you do need to show schematics and/or lines of code. But by the end of the day on Friday, February 18, you should hand in a technical progress report that includes all the information mentioned above. As usual, email your project update report to me by the end of the day on Friday. Name it something like week_7_update_project_name. You can email it to me using PDF, DOCX, or ODT (libreoffice) formats.

Week 8 Discussion: Continue with Biosignals

This week we will continue to look at sensors used with biosignals, and take an introduction to bioinstrumentation systems. The area of bioinstrumentation could easily take two semesters to really cover in depth. But in the time we have, we will look at common human muscle and neve structure, and the kinds of signals one can get off of them. Various signal conditioning strategies that are different from non-biosignal systems will need to be used for several reasons. The techniques used here are useful in other kinds of systems as well, especially those that have low signal-to-noise ratios, and are characterized by sensor systems that have relatively high impedances. We will also look at strategies for non-invasively connecting to a person's body, and the kinds of electrodes that are used for this purpose. Many suitable electrodes you can make yourself!

Week 8 Project:

This week you should be actively building and starting to turn on your sensor based system. If you are building hardware and if the performance of your system is limited by circuit noise similar to what we covered in our general noise model, then you will probably want to make a printed circuit board rather than to use something like a prototyping board that uses plug-in wires which can be very noisy. If your system uses software of any kind, then this week you would be well into writing it and getting it to work. Ideally, this week you should have enough of your system designed and built so that you can start to see data from the sensors, and can start to characterize that data in order to compare it to your original design performance goals. As usual during our scheduled time on Friday your group needs to give a progress update. Remember that this is a good time to ask questions to the entire class to share ideas and solve technical problems. Email your project update report to me by the end of the day on Friday, February 25. Name it something like week_8_update_project_name. You can email it to me using PDF, DOCX, or ODT (libreoffice) formats. Remember that information about how to write technical progress reports can be found here.

Week 9 Discussion: Sensors, Power, Applications and the Future

We will finish up the discussion of Sensor Based Systems by looking at emerging applications including energy harvesting, examples, opportunity areas, and other factors that you will want to take into account as you design sensor based systems for the future. We also can use this week to talk about any other sensor related topic you want, so bring your ideas with you!

Week 9 Homework Assignment:

Note this week there is a homework assignment! The homework assignment should only take you about 1 or 2 hours to do, and is a good way to summarize many of the points we have covered in the course. Here are the instructions for completing the homework assignment.

Get a copy of the homework assignment by clinking on one of the options here. There is a version of the homework in Microsoft Word, another in LibreOffice, and also one in PDF. You can put your answers right on the copy you download and email it to me.
Please do your work on this assignment individually. This answers on this homework should be your own work.
Be sure to show how you arrived at your answers. To get full credit on the answers you need to completely show your work.
You have the full week to complete this assignment. Please email it to me by the end of the day on Monday, March 7, 2022.
Be sure to put your name on the homework. When you email it to me, give the file a name like "your_name_ii2302_homework" so that I can be sure it is yours.

Week 9 Project:

This week you should have your sensor based system well on its way to working. If you need a circuit board be sure to have it done and ready to mill out this week. This week you should be able to start testing and measuring your system so that you can fully characterize it. You want to understand exactly what the system properties are with respect to accuracy, precision, dynamic range, resolution, and noise. You also will want to understand how well your system matches your original estimate for system performance. It may not match your estimates exactly. There is nothing wrong with that because real circuits often don't work exactly as the math or simulations say they should. Circuits that you have built are real. Simulations are not. But having said that, you should be able to explain why the differences exist and how you might be able to improve the system. As you have for the other weeks, email your project update report to me by the end of the day on Friday, March 4. Name it something like week_9_update_project_name. You can email it to me using PDF, DOCX, or ODT (libreoffice) formats.

Week 10:

There are no class sessions scheduled for this week, so a suggestion would be to spend the time working on your projects. At this point in time you should have your sensor based system completely working, and you should be collecting data that fully characterizes the system in a quantitative way. This includes things like accuracy, precisions, resolution, dynamic range and an analysis of noise that affects the system. Presentation and demo of your work will take place next week.

Information about the II2302 Presentation and Demo

The project presentation plus demonstration of complete parts of the project will take place on Wednesday, March 16, 2022 in the Mentorspace starting at 14:00. The presentation covers the entire project to the extent that your group has completed it. It should include the following points:

A description of the system and application you have decided to implement. Be sure to explain why the application is useful, and how it uses sensors.
A description of the software and hardware architecture you have used to implement your project. This description should have block diagrams of your system, hardware schematics, software flowcharts and listings, and other information about how you implemented your system and application.
Data that shows how your design performs, especially with regard to the sensors.
Interpret your data and describe how well the application works.

The entire presentation should take about 10 to 15 minutes. Following that will be a demonstration of as much of the product as you have completed. Everyone in your project group needs to be there to give part of the presentation. A good suggestion would be to have a general presentation of the product similar to what you have done in previous weeks followed by one or two slides from each group member that can be used to help describe what they did in the project.

Final Written Report:

You also will need to hand in a full report about your project. Here is a list of what the report should include. The report is similar to the presentation, but should go into much more detail. Also note the instructions for item #5 below which needs to be done individually.

A description of the system and application you have decided to implement. Be sure to explain why the application is useful, and how it uses sensors.
A complete description of the software and hardware architecture you have used to implement your project. This description should have block diagrams of your system, hardware schematics, software flowcharts and listings, and other information about how you implemented your system and application.
Data that shows how your design performs, especially with respect to the sensors.
A description of who in your project team did what tasks during the project.
Conclusions about your design. Interpret your data and describe how well the application works. Explain performance, especially in terms of the sensors, but also with respect to power consumption, and estimate of cost. Critique your design. Do you think people would buy it? How much money might they spend to buy it? How would you improve your design? What would you do differently if you were to re-design it? And finally, describe what you have discovered in the class. Definitely with respect to sensors and technology, but also with regard to anything else that you discovered while taking the class and doing the project. What have you discovered?

Each project group will hand in one final written report at the end of the term. It is due before the end of the day on Monday, March 21, 2022. In the list above, items 1 through 4 should be included in the final report as a group. However, each group member should write item 5 individually and either attach it to the report or email it to me separately. In other words, there will be a conclusions section for every person in your group. This is a chance to think analytically and critically about what you did, what your results were, and how you would exploit or improve on them. Be detailed in your comments and really think in terms of the design issues that concern sensors and applications that use sensors. Don't forget to put your name on your individual conclusion section.

Please email your group report and individual conclusions to me on or before the due date. Your writeup can be in any of the following formats: PDF, DOC, DOCX or ODT.