My Doctoral Thesis defense is approaching... it will be on Tuesday, June 17th 2014, 10:00 at PA-3 (Level -1 at IST's Math Building).
Scalable and secure RFID data discovery
The combined use of Enterprise Resources Planning (ERP) and Supply Chain Management (SCM) systems has greatly improved the efficiency of supply chains. Further improvements require a deeper connection between the virtual and physical worlds. Automatic Identification technologies, like radio-frequency identification (RFID), allow identification data about tagged physical objects to be collected continuously by readers deployed across locations in the supply chain. This data is stored and managed using traceability systems to allow efficient answers to traceability queries, like Track and Trace.
A practical traceability system should address both the scale and data visibility problems. The system should perform adequately for the number of physical objects flowing in the supply chain; and it should protect the sensitive business data from unauthorized access.
The original contributions of this dissertation are a quantitative assessment framework, called TrakChain, that compares traceability systems for given supply chain scenarios; and a supply chain authorization language that can be used to define and enforce data visibility restriction policies. The results are illustrated with examples from several industries and a case study in the Pharmaceutical supply chain.
--
Descoberta de dados RFID de forma escalável e segura
Os sistemas ERP (Enterprise Resources Planning) e SCM (Supply Chain Management) trouxeram grandes melhorias ao funcionamento das cadeias de fornecimento. Para continuar a melhorar é necessária uma maior ligação entre os mundos virtual e físico. Aqui, as tecnologias de identificação automática por rádio-frequência (RFID), permitem que dados de identificação de objectos físicos etiquetados possam ser recolhidos continuamente por leitores instalados em localizações relevantes (fábricas, armazéns, centros de distribuição, etc). Os dados são guardados e geridos por sistemas de rastreabilidade que permitem dar resposta a interrogações, tais como Localizar e Rastrear.
Um sistema de rastreabilidade prático deve tratar os problemas da escalabilidade e da visibilidade dos dados. O sistema deve ter um desempenho adequado ao número de objectos físicos que circulam na cadeia de fornecimento; e deve proteger os dados de negócio de acessos não autorizados.
As contribuições originais desta dissertação são as ferramentas de avaliação quantitativa, que permitem a comparação de sistemas de rastreabilidade para diversos cenários de cadeias de fornecimento; e uma linguagem de autorização que pode ser usada para definir e aplicar políticas de restrição da visibilidade de dados. Os resultados são ilustrados com exemplos de várias indústrias e com um caso de estudo da indústria Farmacêutica.
Showing posts with label PhD. Show all posts
Showing posts with label PhD. Show all posts
2014-06-03
2009-10-23
Blackboard systems
Blackboards are a problem solving framework. A blackboard is not used to "search for a solution" but to "assemble knowledge".
A blackboard framework provides:
- reasoning flexibility (backward-chaining, forward-chaining, etc);
- participation of varied sources of knowledge;
- opportunistic contributions;
- incremental solutions;
- interleaving of different forms of reasoning.
Problem Solving:
- divide the problem into loosely coupled subtasks (areas of specialization);
- the solution space is divided in regions (levels) containing partial or intermediate solutions;
- the information in each region is globally accessible on the blackboard, making it a medium of interaction between the knowledge sources;
- the decision to employ a particular knowledge source is made dynamically using the current solution state;
- caveat: the problem partitioning in subproblems makes a great difference in clarity, speed, resources, and the ability to find a solution!
Read more about this subject in Blackboard Systems by Englemore and Morgan.
A blackboard framework provides:
- reasoning flexibility (backward-chaining, forward-chaining, etc);
- participation of varied sources of knowledge;
- opportunistic contributions;
- incremental solutions;
- interleaving of different forms of reasoning.
Problem Solving:
- divide the problem into loosely coupled subtasks (areas of specialization);
- the solution space is divided in regions (levels) containing partial or intermediate solutions;
- the information in each region is globally accessible on the blackboard, making it a medium of interaction between the knowledge sources;
- the decision to employ a particular knowledge source is made dynamically using the current solution state;
- caveat: the problem partitioning in subproblems makes a great difference in clarity, speed, resources, and the ability to find a solution!
Read more about this subject in Blackboard Systems by Englemore and Morgan.
2009-10-01
Novidades nos Doutoramentos em Portugal / News about PhDs in Portugal
Por outro lado, importa, ainda, introduzir no sistema de ensino superior nacional a possibilidade de realização do doutoramento com base na compilação de um conjunto de publicações científicas anteriores, publicadas em revistas com comités de selecção de reconhecido mérito, prática hoje internacionalmente utilizada.
Neste sentido, e em condições de exigência equivalentes à da modalidade de doutoramento com base numa tese, introduz-se a possibilidade de obtenção do grau de doutor com base no reconhecimento e análise de obras, e com base em trabalhos científicos publicados.
/ This post is bilingual: Portuguese (Portugal) and English (U.S.) /
From now on, a Doctoral degree in Portugal can be obtained with a collection of published articles instead of the mandatory dissertation.
2009-07-15
Time to dive
You should focus as soon as possible, and "diving into the ocean" is the best way to do it!SS
Hypothesis
A hypothesis consists either of a suggested explanation for an observable phenomenon or of a reasoned proposal predicting a possible causal correlation among multiple phenomena.
(...)
The scientific method requires that one can test a scientific hypothesis. Scientists generally base such hypotheses on previous observations or on extensions of scientific theories.
(...)
A hypothesis is never to be stated as a question, but always as a statement with an explanation following it. It is not to be a question because it states what the experimenter thinks will occur. Hypotheses are usually written in the "if-then form": If X, then Y.
Source: Wikipedia
(...)
The scientific method requires that one can test a scientific hypothesis. Scientists generally base such hypotheses on previous observations or on extensions of scientific theories.
(...)
A hypothesis is never to be stated as a question, but always as a statement with an explanation following it. It is not to be a question because it states what the experimenter thinks will occur. Hypotheses are usually written in the "if-then form": If X, then Y.
Source: Wikipedia
2009-06-23
Work proposal layout
This is an interesting way to present a work plan in a dissertation:
"
Proposed work item
Completed work item
On-going work item
Goal, related publications
Hypothesis - method, expected outcome, intuition
"
--
Credit for this idea goes to Mihai Rotaru's PhD thesis proposal.
And to Joana for bringing my attention to it.
"
Proposed work item
Completed work item
On-going work item
Goal, related publications
Hypothesis - method, expected outcome, intuition
"
--
Credit for this idea goes to Mihai Rotaru's PhD thesis proposal.
And to Joana for bringing my attention to it.
2009-06-17
Writing tip: avoid contractions
I received a good tip about thesis-quality writing: avoid contractions like "isn't" or "don't" and use the complete "is not" and "do not". It makes the text more formal.
Do you suggest any other English writing tips? Feel free to leave a comment.
Do you suggest any other English writing tips? Feel free to leave a comment.
2009-03-17
Tweeting live from João Barreto's PhD defense
Tweeting live from João Barreto's PhD defense. Good luck João!
The committee members are putting on their robes. José Carlos Monteiro is wearing the MIT robes, hat included.
José Alves Marques is presiding. The defense has started. First the 30 minute presentation.
João is explaining the problem. The presentation graphics look nice.
Thesis goals were rapid commitment, and minimal aborted work.
contributions a) commitment protocol, b) exploitation of non-replicas for improved consistency, c) redundancy elimination in storage and propagation of state-based updates.
João is explaining the commitment protocol based on epidemic voting (VVWV).
Moving on to b) exploitation of non-replicas. It's an interesting idea. They carry only the voting information.
Getting to c) redundancy elimination. João is placing his related publications in each slide's footer.
João has been presenting his results using graphs. It's always good to see hard data.
1 minute conclusions slide. Maybe too fast? Now all 9 publications are on display in the final slide.
Marc Shapiro is now asking questions to João.
Discussing whether application semantics should or shouldn't be used.
Carlos Moreno (DI/Univ. Minho) made some good overall comments about the thesis. Now going chapter by chapter.
Carlos Moreno is asking some questions to João. Some of the issues were out of the thesis' scope
Q&A with Carlos Moreno continues... Now discussing the possibility of applying the results to sensor networks.
Carlos Moreno asking a question about the sampling (Are 5 runs enough to average?). Now moving to final question.
José Carlos Monteiro (JCM) is now addressing João. Particularly liked Chapter 2. It explains all the required concepts.
JCM found a gap: algorithms are for limited devices but there are no figures for resource consumption of algorithms.
José Alves Marques (JAM) is talking now. Lot of effort in the thesis. Chapter 2 is very good although a glossary missing.
JAM is asking what is the rationale of the thesis - or is the thesis just a sum of 3 parts?
Paulo Ferreira is the final speaker. It's an important day for João - he's wearing a tie!
João started his work in early 2000s. Research started with MS ad-hocFS. He has a good publication record.
Paulo Ferreira is giving some nice questions to João. It's a good thing to have a friendly adviser.
The defense is over. We're now waiting for the announcement of the new Doctor...
Congratulations Doctor João Barreto! It was a pleasure being here. Now, back to Distributed Systems course work and my own PhD.
The committee members are putting on their robes. José Carlos Monteiro is wearing the MIT robes, hat included.
José Alves Marques is presiding. The defense has started. First the 30 minute presentation.
João is explaining the problem. The presentation graphics look nice.
Thesis goals were rapid commitment, and minimal aborted work.
contributions a) commitment protocol, b) exploitation of non-replicas for improved consistency, c) redundancy elimination in storage and propagation of state-based updates.
João is explaining the commitment protocol based on epidemic voting (VVWV).
Moving on to b) exploitation of non-replicas. It's an interesting idea. They carry only the voting information.
Getting to c) redundancy elimination. João is placing his related publications in each slide's footer.
João has been presenting his results using graphs. It's always good to see hard data.
1 minute conclusions slide. Maybe too fast? Now all 9 publications are on display in the final slide.
Marc Shapiro is now asking questions to João.
Discussing whether application semantics should or shouldn't be used.
Carlos Moreno (DI/Univ. Minho) made some good overall comments about the thesis. Now going chapter by chapter.
Carlos Moreno is asking some questions to João. Some of the issues were out of the thesis' scope
Q&A with Carlos Moreno continues... Now discussing the possibility of applying the results to sensor networks.
Carlos Moreno asking a question about the sampling (Are 5 runs enough to average?). Now moving to final question.
José Carlos Monteiro (JCM) is now addressing João. Particularly liked Chapter 2. It explains all the required concepts.
JCM found a gap: algorithms are for limited devices but there are no figures for resource consumption of algorithms.
José Alves Marques (JAM) is talking now. Lot of effort in the thesis. Chapter 2 is very good although a glossary missing.
JAM is asking what is the rationale of the thesis - or is the thesis just a sum of 3 parts?
Paulo Ferreira is the final speaker. It's an important day for João - he's wearing a tie!
João started his work in early 2000s. Research started with MS ad-hocFS. He has a good publication record.
Paulo Ferreira is giving some nice questions to João. It's a good thing to have a friendly adviser.
The defense is over. We're now waiting for the announcement of the new Doctor...
Congratulations Doctor João Barreto! It was a pleasure being here. Now, back to Distributed Systems course work and my own PhD.
2008-10-06
More assertive talks
How to make my talks more effective? I got some top-notch advice from people I'm working with right now.
Don't just present facts. Show that you applied your wisdom in preparing the talk. Second, begin with a short introduction, then introduce your hypothesis/conclusion to capture the audience's attention! If you leave it to the end, people are just spectators during most of the talk. Next, go through the hypothesis/conclusion items, one by one, justifying them (the whys). At the end, summarize what the talk was about.
- The presentation style should pull audience from their collars; punch-line first; i.e. tell the story from end, twist, and then to beginning;
- Never repeat outside information;
- Challenge status quo;
- Make "Wall Street Journal" headlines;
- Try to always have demos or connections to physical world - it draws people's attention, particularly if they don't care that much about the subject of the talk.
The difference between the top and the rest of the world is the arrogance of people. They may be wrong 50% of the times, but they're right the other 50%. Self-confidence is a requirement to move forward. That's how change can happen!
This is particularly true when trying to promote synergy between higher education and industry:
Don't just present facts. Show that you applied your wisdom in preparing the talk. Second, begin with a short introduction, then introduce your hypothesis/conclusion to capture the audience's attention! If you leave it to the end, people are just spectators during most of the talk. Next, go through the hypothesis/conclusion items, one by one, justifying them (the whys). At the end, summarize what the talk was about.
- The presentation style should pull audience from their collars; punch-line first; i.e. tell the story from end, twist, and then to beginning;
- Never repeat outside information;
- Challenge status quo;
- Make "Wall Street Journal" headlines;
- Try to always have demos or connections to physical world - it draws people's attention, particularly if they don't care that much about the subject of the talk.
The difference between the top and the rest of the world is the arrogance of people. They may be wrong 50% of the times, but they're right the other 50%. Self-confidence is a requirement to move forward. That's how change can happen!
This is particularly true when trying to promote synergy between higher education and industry:
2008-06-16
A new adventure starts...
No, I'm not talking about the latest Indiana Jones movie... I'm talking about my PhD! After the prologue, it's about to really begin! I've got so much work ahead, but I'm looking forward to it!
It wouldn't be possible without some very special persons... they know who they are ;-)
It wouldn't be possible without some very special persons... they know who they are ;-)
2008-02-04
PhD Position Paper - v1.0
Today I finished the first complete version of my position paper. One month late according to my original planning.
The title is "Computer science research for the Internet of Things". The paper is about the forthcoming widespread use of smart things, like RFID tags and sensors, along with omnipresent wireless networks, that will create an Internet of Things (IoT), where most everyday objects will be interconnected and part of an universal-purpose system akin to the Internet.
This position paper looks at the IoT in a computer science research perspective. It identifies the main challenges to address and points to a set of possible research paths. It also provides some insight into what will change in the way information systems are designed and used in the IoT era.
The paper is still not ready for publishing, but it's ready to be sent to my Advisor so we can discuss the written ideas.
In short, the PhD prologue still isn't over, but I'm making progress...
All the best!
The title is "Computer science research for the Internet of Things". The paper is about the forthcoming widespread use of smart things, like RFID tags and sensors, along with omnipresent wireless networks, that will create an Internet of Things (IoT), where most everyday objects will be interconnected and part of an universal-purpose system akin to the Internet.
This position paper looks at the IoT in a computer science research perspective. It identifies the main challenges to address and points to a set of possible research paths. It also provides some insight into what will change in the way information systems are designed and used in the IoT era.
The paper is still not ready for publishing, but it's ready to be sent to my Advisor so we can discuss the written ideas.
In short, the PhD prologue still isn't over, but I'm making progress...
All the best!
2007-09-27
Searching digital libraries
I'm currently looking for research papers on the top scientific digital libraries. For computer science, they are ACM's and IEEE's. Also interesting are Google Scholar and CiteSeer.
My search strategy was to build a table where the rows are the search terms (ex. RFID, Auto-ID, ...) and the columns are the libraries or search engines (ex. ACM, IEEE, Google Scholar, CiteSeer, ...).
I then went through all the table cells, doing the searches and looking at the results.I read the paper's titles and abstracts and selected those more relevant for my current interests. This can be very time consuming. For a 4x4 table, I downloaded a total of 200 papers.
For all downloaded papers, I also stored bibliographic information, preferably in BibTeX format, or at least, the paper's URL location. This will be very handy when I'm writing my own papers.
I'll take note of the date when the search was performed. Then, when I want to update the search, I can then specify that I only want papers published after my last search (this option is usually available as an advanced search parameter).
Another interesting information to retrieve is to identify which conferences are more relevant to my current work, as these can be good publication targets. I put them in my Google conference calendar to keep track of the submission dates.
My search strategy was to build a table where the rows are the search terms (ex. RFID, Auto-ID, ...) and the columns are the libraries or search engines (ex. ACM, IEEE, Google Scholar, CiteSeer, ...).
I then went through all the table cells, doing the searches and looking at the results.I read the paper's titles and abstracts and selected those more relevant for my current interests. This can be very time consuming. For a 4x4 table, I downloaded a total of 200 papers.
For all downloaded papers, I also stored bibliographic information, preferably in BibTeX format, or at least, the paper's URL location. This will be very handy when I'm writing my own papers.
I'll take note of the date when the search was performed. Then, when I want to update the search, I can then specify that I only want papers published after my last search (this option is usually available as an advanced search parameter).
Another interesting information to retrieve is to identify which conferences are more relevant to my current work, as these can be good publication targets. I put them in my Google conference calendar to keep track of the submission dates.
2007-09-10
Big Picture day
At this stage of my PhD work I want to keep an open mind to new ideas, however I don't want to lose my focus...
How can I cope with both these seemingly contradictory goals?
I've tried several things in the past, but now I'm trying to have a "Big Picture day" in the week. I use this work day to keep track of the cool and interesting science and technology stuff out there: I read papers, news, mails, web sites, etc.
If I make a new connection with my work, I'll write it down and look it up later.
In the remaining days of the week I refrain from looking at stuff, to avoid distraction.
How can I cope with both these seemingly contradictory goals?
I've tried several things in the past, but now I'm trying to have a "Big Picture day" in the week. I use this work day to keep track of the cool and interesting science and technology stuff out there: I read papers, news, mails, web sites, etc.
If I make a new connection with my work, I'll write it down and look it up later.
In the remaining days of the week I refrain from looking at stuff, to avoid distraction.
2007-09-03
Back to work: PhD Prologue
I'm coming back to work after a fantastic holiday with the people I love the most.
Now I've got my "batteries" fully recharged to start a new challenge: my PhD.
Before I enroll in the actual course, I'm doing something I named PhD prologue. I've been doing it since January 2007, but now it's focusing and intensifying.
During this prologue I spoke with several researchers and senior colleagues, looking for interesting subjects for my PhD.
I'm currently directing my preliminary research to the Internet of Things, that aims to leverage the power of the Internet and of the upcoming RFID technology.
RFID stands for "Radio Frequency IDentification" and is a technology that allows contact-less reading of data tags or sensors. In a world where each object is not only classified (with a bar code), but also identified (with a RFID tag), imagine the new problems (and opportunities) that are coming just around the corner...
The goals of the PhD prologue are:
1 – Delineate the research area
2 – Identify the players (industry and academia)
3 – Identify the main problems to address
I intend to write about these in a position paper. Then, the PhD can start!
Now I've got my "batteries" fully recharged to start a new challenge: my PhD.
Before I enroll in the actual course, I'm doing something I named PhD prologue. I've been doing it since January 2007, but now it's focusing and intensifying.
During this prologue I spoke with several researchers and senior colleagues, looking for interesting subjects for my PhD.
I'm currently directing my preliminary research to the Internet of Things, that aims to leverage the power of the Internet and of the upcoming RFID technology.
RFID stands for "Radio Frequency IDentification" and is a technology that allows contact-less reading of data tags or sensors. In a world where each object is not only classified (with a bar code), but also identified (with a RFID tag), imagine the new problems (and opportunities) that are coming just around the corner...
The goals of the PhD prologue are:
1 – Delineate the research area
2 – Identify the players (industry and academia)
3 – Identify the main problems to address
I intend to write about these in a position paper. Then, the PhD can start!
Subscribe to:
Posts (Atom)