Mark Post
I build robots!
Activity
-
Hi Kabeer. This is indeed a great application. For self-driving cars it has been shown that far more cars can be accommodated on a highway if they work as a communicating swarm - the whole highway can respond as a unit to any changes. The main problem right now is how to integrate this many cars with the highway itself...and even more of a problem, how to...
-
Mark Post replied to Sainey Danso
Hi Sainey. Good observation - in general the more sensors the better as they can cover each others' weaknesses...
-
Mark Post replied to ismaeel khan
Hi Ismaeel. VR is a great tool for many telepresence situations - swarms are especially challenging but especially good for hazards also. Joysticks do indeed usually get used in the end because they are so intuitive and accurate, but there are alternatives too.
-
Mark Post replied to Keith McGraw
Keith, good topics that you mention here - there is always a slight delay in communications (perhaps more obvious now because of the widespread use of tele-meetings these days) and systems like ROS (very useful as you say!) have to include timestamps to correct for sensor data delays even if they are on the same robot.
-
Mark Post replied to EMMANUEL DEBRAH
Hi Emmanuel. You have some very good ideas here. Compression of data from distant sources is very valuable (think .zip files) for saving bandwidth as you mention, and also some new approaches actually identify which data is more important (more interesting for the mission or application) and send them first, sending less important data only if requested.
-
Hi Joseph. Very true - in a lot of cases rather than true telepresence, data is just received from the rovers some time later without concern for synchronicity of operation.
-
Hi Peter. I'm glad to hear that your work is leading you to more interesting challenges - automation of semiconductor doping processes sounds very interesting! Indeed the next big advances will probably have to be in AI and autonomy as humans still have to do most of the complex work as robots (especially in space) are very simplistic.
-
Mark Post replied to Sainey Danso
Hi Sainey. Very good point - in a sense, telepresence is only as good as the infrastructure it can use...
-
Mark Post replied to Joseph Kininmonth
Hi Joseph. This is a good idea - already some telepresence technologies are used (e.g. cameras on the end of manipulator arms) so that astronauts do not need to space walk to do external repairs. Robonaut 2 for example could be telepresence controlled in a space walk. (https://robonaut.jsc.nasa.gov/R2/)
-
Mark Post replied to Ben Porter
Hi Ben. Very true, we are always in need of more bandwidth, especially for science instruments and Earth observation that can produce many gigabytes of data quickly. To some extent this is seen in the 5G cellular technologies that operate at higher frequencies, but are much shorter range as a result, so a relay station is needed on almost every street...
-
Mark Post replied to Peter Walsh
Hi Peter. Not at all, this is a very good idea, and is the principle behind NASA's deep space network among others - spacecraft orbiting Mars and Earth act as "relay points" to ensure communications gets through (though latency is inevitably added too).
-
Mark Post replied to Tony Whicker
Hi Tony. This is a great idea! Telepresence broadcasting has been done from the International Space Station and from sites on Earth, and as communication technology increases will likely see more applications. The one problem is that usually only one person can control (outgoing information and commands) the robot at once. But maybe in future there will be...
-
Mark Post replied to Ormila Azim
Good points, Ormila. There are a lot of interesting applications that telepresence robots are now being used in, and they will undoubtedly keep growing.
-
Hi Ormila. Great description. And we will probably see a lot more of these capabilities on "social" robots in the near future.
-
Hi Abdullah. This is a good point actually - the use of telepresent-like technologies for improving security. Also, we are now looking at autonomous vehicles for mapping the sea floor because it is very difficult to communicate underwater without long tethers...but if this problem can be solved it would be outstanding for exploring the sea floor.
-
Hi Ormila. We could potentially deal with large objects this way - at the risk of creating a lot of debris, which is why it's kind of a last-resort option. Other possibilities include attaching a thruster (a popular solution for de-orbiting debris also) or firing a laser at it to push it with the resulting superheated gases.
-
Hi Abdullah. This is a good point - where do we keep the consumables for humans? We already have semi-automated supply vehicles for the ISS, perhaps we will have a "wagon train" going between Earth and Mars someday.
With regard to velocities, it's not difficult to cool things in deep space even at high velocity as there is no friction (though if you have a... -
Hi Emmanuel. Indeed backups are nearly always used for all systems on a spacecraft - at least three are necessary for triple modular redundancy (TMR)
-
Hi Tony. Yes, this is a really good idea because there is a layer of Venus' atmosphere that is actually about the same pressure and temperature as Earth. Balloons can float around the planet in this layer and be very comfortable compared to a lander.
-
Hi Emmanuel. You make a good observation here - we do need better battery technology and methods to maintain them and monitor them better, while in the difficult environment of space that makes conventional batteries hard to use.
-
Hi Graham. Well said, and we need to remember that these complex devices only work at all because scientists and engineers have been working on them for decades, and even longer. We already can accomplish some pretty miraculous things in this context.
-
Hi Ben. Great observations on this topic! Indeed it is a problem that we can't "joystick" these missions. AI and better autonomy would be a great help but we are not yet at the point where we can rely on autonomy alone for a complex mission. Once we have smarter robots on Earth we will make progress in having spacecraft take better care of themselves (e.g....
-
Hi William. Indeed temperature and light are key problems - if we had a way to generate unlimited power both can be addressed better, and to some extent we do with Radioisotope Thermoelectric Generators (RTGs). The problem is they are dangerous and avoided whenever possible as they contain radioactive materials.
-
Mark Post replied to Keith McGraw
Hi Keith, great mentions of bird strike resilience and abrasion resistance in applications on Earth (I have heard apocryphal stories of using frozen chickens for testing cockpit windows...) I suppose the problem in space is that in many cases we don't have systems strong enough to handle any sort of significant impact, and it may stay that way until we have...
-
Mark Post replied to Peter Walsh
Hi Peter. Great to hear that you are studying Mechatronics - that's one of the most vital areas for space hardware development. The main difference is that in space the tolerances are that much finer as well as the environmental resilience being more challenging.
-
Mark Post replied to Alwyn watson
Hi Alwyn. Good observation, the most effective radiation shields tend to be layered to work against different kinds of radiation. Also to decrease overall mass layers can be used.
-
Mark Post replied to Alwyn watson
Hi Alwyn - good strategy! For micro-meteoroid impact protection a variety of layered shields have been tested - a good list is at https://hvit.jsc.nasa.gov/shield-development/ and lasers are also being considered (though mounting one on the ISS might make it a better idea as you suggest) - NASA's project at...
-
Mark Post replied to Abdullah Omar
Hi Abdullah. Very true, especially a spacecraft that has to function in space, and then on planets like Venus which are opposites in the kind of environment.
-
Mark Post replied to Tony Whicker
Hi Tony. Good observation. Indeed they can, and sometimes do, use superconducting components in space (e.g. electric thrusters and super-sensitive magnetic sensors). However it is not too common because spacecraft in the sun heat up very quickly, and this stops superconductivity unless good shielding measures are used (e.g. the heat shield on the James Webb...
-
Mark Post replied to Keith McGraw
Keith, thanks for your very insightful comments from your experience. Materials are often not highlighted in engineering applications, but are the largest factor in performance and survivability. Also it's important to note that we do not have unlimited supply of some of the elements and compounds needed for these applications. E.g. the superalloys that the...
-
Mark Post replied to Keith McGraw
Hi Keith. This is a very true and somber message. Whenever we go somewhere new we usually end up contaminating it before we learn to conserve it. Now is the time to learn from the messages of history.
-
Mark Post replied to Ben Porter
Hi Ben. Indeed a danger, and difficult to catch - we don't yet have a net that can slow down and capture the smallest (and potentially most dangerous) objects.
-
Mark Post replied to Peter Walsh
Hi Peter and Keith - definitely the way to go and good points. In the early days of space exploration (Apollo, etc.), rocket bodies were often just left in orbit. Now they must be deorbited specifically after use. SpaceX's automated return/re-use saves a lot of waste from these launches. (ultimately of course the goal is zero waste, which will require...
-
Mark Post replied to Edward Holden
Hi Edward. Thanks for highlighting this! Not sure if this is the one you are referring to but it is on now: https://www.bbc.co.uk/programmes/m000kqm9
-
Hi Sriramsingh. Good catch! Copper does indeed melt at 1084 degrees C. The text should read "Other metals such as Copper will lose strength at even lower temperatures". "Cold-worked copper can be softened at relatively low temperatures (~200 C) because of its low recrystallization temperature." (source: Taubenblat, P. W.; Smith, W. E.; Graviano, A. R. In...
-
Hi Sainey. Yes, but even more so the radiation from the sun that we do not receive on the surface of the Earth.
-
@PeterWalsh Very close to absolute 0. Having worked with thermal vacuum chambers I remember that these cryo pumps can produce pressures very close to hard vacuum (below 1e-6 torr or 1e-4 Pa ~ish). Very useful for testing hardware in vacuum before launching it.
-
Hi Tony. Indeed a good point, and also how they operate in a vacuum (anything that vaporizes tends to coat nearby surfaces causing very unpleasant effects).
-
Mark Post replied to Ben Porter
Hi Ben. Obscuring senses is a very good example of a hazard - it is easy to be harmed if you cannot avoid things in the environment. For this reason, robots and vehicles need a lot of sensors in case some do not work.
-
Mark Post replied to Katy MacDonald
Hi Katy. Very good observation, "Hazardous" in fact has no meaning unless there is some context in an object it is hazardous to!
-
Mark Post replied to William Jeffery
Hi William. You make a good point - lack of sustenance is a hazard, whether it is air, food, water, or sunlight even.
-
Mark Post replied to Joseph Kininmonth
Hi Joseph, yes - maintaining our life in most cases, and sometimes the "life" of the technologies that we create.
-
Mark Post replied to Walter Romero
Hi Walter. Indeed geography is a big part - and if you consider acid rain then Venus is a great example...
-
Mark Post replied to Edward Holden
Hi Edward. Good point - you bring in the factor of time and it is indeed usually a matter of how long an entity can tolerate certain environmental factors
-
Mark Post replied to Abdullah Omar
Hi Abdullah. Indeed this is the basic definition of "hazardous", although in some cases we put the robots in place of ourselves!
-
Hi Jennifer. Yes, the better the sensors are the harder reliability and latency become - but a human operator then has a much better understanding of the environment too, so it is worthwhile. Regarding factories, a lot of large-scale warehousing is now fully automated by robots, making it much more convenient and fast for workers to work.
-
Mark Post replied to Kaat Lambrecht
Hi Kaat. Yes, certainly swarm robots are good for accessing difficult places - also if only simple sensors are needed very good coverage is possible.
-
Mark Post replied to Kirsten Downie
Hi Kirsten. Indeed - and this is why the robot should be clever enough to be aware of its surroundings and "help" the telepresent user by combining information and making suggestions.
-
Hi Sabrina. Some really good ideas in what you have said. Typically the more sensors a robot has, the better it can work because there is more useful information at hand (and the better telepresence can work also). This "more is better" philosophy also increases complexity, which is why complexity management is a key driver in modern robotics.
-
Mark Post replied to Ben Murphy
Hi Ben. These are a lot of good observations. Humans are most used to their own sensing set (vision, hearing, smell, touch, taste) and removing or altering them in a telepresence situation tends to throw us off. It does indeed work much better if the robots do the sensing and decision making locally. The challenge now is making them clever enough to do the...
-
Mark Post replied to Danny Rashley
Hi Danny. Indeed swarms are great for covering ground fast. A lot of current research involves connecting them together wirelessly to work as a unified whole, or connecting them physically to form morphable robots able to reach difficult places.
-
Mark Post replied to Renata Loberto
Hi Renata. This is true in some situations, but it really depends on the challenge at hand. While just being able to wait (if data is not time-critical) can solve latency, most communications systems include error correction codes that can improve reliability or detect if an error has occurred. You are very correct though that the combination of high...
-
Mark Post replied to Lu Loughlin
Hi Lu. This is a good example of a similar situation - competitive online games have a lot of clever communications features to ensure that critical data gets sent as close to real-time as possible, and of course the best gaming controllers and monitors must have no lag at all!
-
Hi Richard. In fact, this is already done in most planetary rovers. Spirit and Curiosity were not very autonomous, but they did have navigation cameras and an "emergency stop" function in case they get too close to a cliff or rock. Also, different information is sent (to some degree) on different frequencies to orbiting spacecraft - typically high-bandwidth...
-
Mark Post replied to Kirsten Downie
Hi Kirsten. Indeed, space roboticists must have very good patience!
-
Mark Post replied to Kaat Lambrecht
Hi Kaat. You make reasonable points - if there are no obstacles than the importance of latency (and reliability) decreases. However this is only because there is no penalty to re-trying and waiting - if your luck is bad you may never get a completely error-free communication, and so it is still worth placing bounds on how long or bad communication is.
-
Mark Post replied to Tess Manson-Whitton
Hi Tess. Great point you have made. It is only because of our "telepresence" capability over the internet and with mobile devices that our society continues to function even in the present pandemic.
-
Hi Pranav. Perhaps this would have been a good design tradeoff - in fact there is some discussion on whether Jupiter's massive magnetic fields could power a spacecraft themselves. Nuclear power is generally avoided because of the much higher design complexity, mass, and the safety risks to humans in the event of a failed launch, unless there is no...
-
Hi Pranav. A "booster" like this would be a good idea. Perhaps in future missions they will design the lander to jump out of a bad situation if one occurs.
-
Hi Mary. This would be great indeed. It is a matter of the tradeoffs they decided to make for battery size versus solar power and they didn't expect to mess up the landing to that extent.
-
Hi Erica. Indeed this would be ideal if we could do it.
-
Mark Post replied to Paul Mitchell
Hi Paul. Well, we could retrieve and recycle space junk - there are a lot of valuable materials in them after all. However the problem right now is not the recycling but the retrieval.
-
Mark Post replied to Pranav Gajula
Hi Pranav. Actually they do use silicon electronics in deep space but the problem is that it has to be redesigned to have larger band gaps and tolerances to overcome radiation and temperature extremes. And often heaters must be used that consume power - hence you are correct in the end, on a deep-space mission that does not have unlimited power we will need...
-
Mark Post replied to Kaat Lambrecht
Yes, it's called "dynamic range" of conditions - it's not as hard to protect against one extreme as to protect against extremes in both high and low range.
-
Mark Post replied to Pranav Gajula
Hi Pranav. It's not specifically hard to reach that speed, but the faster you go the more energy you need to put in. Spacecraft carry limited fuel so the faster they have to go the heavier they get. That's why a slingshot manoeuver was necessary to gather extra speed.
-
Mark Post replied to Ondrej Duda
Very true, Ondrej. Humans tend to change their environment (sometimes making it hazardous...)
-
Mark Post replied to Renata Loberto
Hi Renata. The idea was that they wanted to avoid leaving any space junk (it's enough of a problem around Earth as it is) so whenever a spacecraft reaches the end of its useful life (before it becomes nonfunctional) the standard practice is to crash it into a planet or whatever is nearby to prevent proliferation of space junk.
-
Mark Post replied to Kaat Lambrecht
Hi Kaat. That is indeed the concern now - and why cleaning up space is now the highest priority of some space organizations.
-
Mark Post replied to Ondrej Duda
Hi Ondrej. This is a good point - environments can become hazardous for a variety of reasons
-
Mark Post replied to Erica Obiefuna
Hi Erica. Good summary!
-
Excellent comment, Richard! In fact, due to time constraints many missions do not even launch with the final version of the software they will use! The Mars Exploration Rovers received software updates in transit to Mars, and were sent software modifications whenever it was deemed to be needed. Of course they don't want to break functioning software, so...
-
Hi Lizzy. Good point. Beyond the question of telepresence (elsewhere in this MOOC) space communication is a complex discipline in itself. NASA's Deep Space Network is probably the biggest communications asset to exploration of our solar system at present, but ESA's ESTRACK is also quite capable and other facilities are being developed too.
-
Well said, John. Indeed!
-
Excellent point, Richard. The early missions to the planets were focused almost entirely on answering these questions. Until the Russian Mars and American Viking probes, scientists still thought that there might be canals of water on Mars! Even by the time of the Apollo 11 mission, some scientists thought that the Astronauts might sink straight into the...
-
Hi Kirsten. Very astute observation. Robots can have a food/energy crisis too because of limitations on space and mass for battery power. It would have perhaps been better if they had not relied on solar power and just provided a larger storage battery...but that was a calculated design decision. If they'd landed as expected they would have had lots of...
-
Hi Lizzy. Very good summary. And they had neither a precise landing map nor a lot of gravity (and it didn't stick in the end...)
-
Good observations, Jennifer. The terrain is almost impossible to predict so a lot of it comes down to luck. Making Philae mobile would have been a great help in this case - even just giving it enough intelligence to right itself with an arm and crawl out of a shadowed area would help. It's really hard to do this in an unfamiliar environment without creating...
-
Mark Post replied to Z.AHAMED AAKIF
Hi Ahamed. Glad you liked the video. Indeed, we try to make them survive "anything" but because of other constraints (cost, time, mass, complexity, etc.) it is usually hard enough just making them survive the specific challenges of a given mission. Of course, the more challenges a mission has, the harder it is...(but the tougher they are in the end too!).
-
Hi Sabrina. Absolutely right, a large part of the problem is that in the time it takes us to recognize the problem, the problem gets worse. International legislation requiring de-orbiting solutions is only now starting the process of implementation, but by the most reputable accounts, we are already seeing Kessler syndrome happening.
-
Hi Richard. This is an interesting way to look at it - we build something to survive extremes and then have to get rid of it. Great example and yes, recycling buildings is a significant challenge. Also relevant to all the space junk we put up in orbit.
-
Mark Post replied to Lu Loughlin
Hi Lu. This is a good example that is not so obvious - a place you are not familiar with that could cause unexpected collisions. This is a major challenge with sending robots to remote places.
-
Mark Post replied to Z.AHAMED AAKIF
Hi Ahamed. I like your logical way of deconstructing the term. A hazardous environment can contain any "hazard" at all (relative to the thing it is hazardous to).
-
Mark Post replied to Tess Manson-Whitton
Hi Tess. Very good that you mention man-made environments - nuclear reactors or chemical areas are great examples. Not only are they dangerous to us but they are dangerous to machines we make.
-
Mark Post replied to Kirsten Downie
Hi Kirsten. Very true for humans. In more general sense, it could be just a "danger" to anything...
-
Mark Post replied to Kaat Lambrecht
Hi Kaat. Great concise way of putting it!
-
Mark Post replied to Jennifer Wataranan
Hi Jennifer. All good points - I sense a definition forming around "anything that can hurt you..."
-
Hi Ben. Excellent observation! For space technology to last very long times without shutting them down regularly or maintaining them, we will need new breakthroughs in electronics and also mechanical components. Faster space travel would be wonderful (we've all dreamed about warp drive) but we're a lot closer right now to having a spacecraft last for 50...
-
Hi Megan. Very much so! The materials have to be carefully selected to be strong and tolerant to the hazards expected.
-
Hi Ben. Great set of observations. Comets in particular tend to be rather loosely held together, and easily start to evaporate or disintegrate when they pass through the Solar Wind. So not only is the surface unstable, but also it's very easy to collide with nearby particles (the Hayabusa missions to take samples from comets could easily scoop up fragments,...
-
Hi Sabrina. Absolutely correct. It's hard for us to create technology that can handle places that are so different than on our Earth. We're used to a lot of sunlight and don't have to worry about dodging micrometeoroids, so it's a big challenge to get used to the conditions elsewhere in the Solar System.
-
Mark Post replied to Ben Murphy
Hi Ben. Great observation. This need for larger "dynamic range" in conditions causes a lot of challenges in space engineering. For example, a spacecraft could experience over 100*C temperatures if it is orbiting between the Earth and the Sun, and then -60*C temperatures a few hours later in eclipse on the opposite side of the Earth. Spacecraft often have...
-
Great summary, Dom! It's interesting to note that these measures are mostly necessary because we come from a place with "different" norms. If we somehow evolved on Venus, the 500 degree temperatures might not be a big deal...
-
Mark Post replied to Danny Rashley
Hi Danny. Indeed, the more "hazardous" it is the less can exist there (of course everything is relative).
-
Mark Post replied to Saadi Mohammed Ahmed
Hi Saadi. Good guesses at specific hazards - both can be dangerous in some cases.
-
Mark Post replied to Alan Poole
Hi Alan. This is a good definition - essentially if it can hurt you, it is hazardous (of course that depends on what "you" are...)
-
Hi Richard. Indeed, it's not just "one sense" that must be used, it is many senses and tests together because of the number of problems that could exist.
-
This is a good point - quite often the appearance of the water is misleading (after all, rubbing alcohol looks just like water too!) You have to use your knowledge and scientific tests to determine how safe something is.
-
Hi Lizzy. Absolutely. Although it is more difficult and expensive to find better energy and resource alternatives, we have had a "free ride" for decades on fossil fuels, and the damage is becoming more and more obvious (see all the recent articles published on global warming, glacial melt, animal migration patterns changed, etc.)
-
Hi Dom. Yes indeed! And you have noted that there are different problems in different places, that often require different approaches. Engineering must always consider the context of its effects.
-
Mark Post replied to Spencer Barber
Hi Spencer. Definitely a great use of computing and databases. Some recent work actually built a neural network out of known responses in the human body and then used that neural network to identify responses that would cause certain symptoms. There will be a lot more of this kind of large-data modelling in the near future.
