0:02 it's my great honor and pleasure to
0:07 introduce Professor Pao cap caparo from
0:10 MIT and Professor cap caparo is the
0:14 fourth professor of engineering at MIT
0:16 she is a professor of nuclear science
0:19 and engineering a professor of physics
0:21 and a member of research lab for
0:24 electronics where she leads the quantum
0:27 Engineering Group Professor caparo
0:30 received her PhD from MIT and joined
0:33 Harvard University as a postdoctoral
0:35 associate in The Institute for
0:39 theoretical Atomic molecular and Optical
0:43 physics before joining back before going
0:46 back to MIT as a faculty okay please
0:50 join me to welcome
0:52 Professor thank you and and frma you
0:54 should stop your sharing otherwise I
0:57 cannot start mine thank
1:09 okay can you hear me and uh see my
1:12 slides yes yes thank you
1:14 perfect uh thank you very much for the
1:17 invitation to present and uh also for
1:20 the uh very nice uh scheduling which had
1:22 me just after fre my presentation
1:25 because I'm sort of hopefully continuing
1:27 on the same topics and uh uh with
1:30 somehow a complimentary um view if you
1:34 want uh in which I'm I'm really trying
1:36 uh to sort of tease out uh what is
1:39 quantum in in Quantum sensing if you
1:42 want and uh why we want to do Quantum
1:44 sensing and what type of Advantage we
1:48 can have over a classical sensor uh when
1:51 we are exploiting quantum properties and
1:53 so the the first message that I really
1:55 want to convey is why are we interested
1:58 in using Quantum system uh for sensing
2:01 and this sort of goes back to uh the
2:03 very old thought experiment of shinger
2:06 the the famous shinger cut uh that can
2:09 be um a awake or or sleeping inside the
2:12 box and we don't know uh what it's doing
2:15 until we open the box but really uh when
2:19 we look at a real experiment well uh we
2:21 cannot really a superposition of a
2:23 Quantum cut or of if you want a
2:25 microscopic cut and that's is because
2:28 the cut even inside the box is still
2:31 interacting with its environment any
2:33 molecule of air any photons Inside the
2:36 Box will perur it and immediately
2:38 collapse it into other being awake or or
2:41 or sleeping and this is because Quantum
2:43 system are fragile because they interact
2:46 very strongly uh with uh the external
2:49 perturbation and this is probably bad if
2:51 you're trying to make a quantum computer
2:54 because you would like to keep this a
2:56 very large superp position but they also
2:59 make for very good sensor there a high
3:04 sensitive it uh to um any external field
3:06 and external perturbation makes Quant
3:09 system very good Quantum
3:13 sensor and uh of course the uh the goal
3:16 that we have is to reach sensitivity
3:19 which can be beyond what is possible
3:22 with classical sensor and this is a form
3:24 of quantum Advantage which sort of
3:26 mimics the advantage that you have in
3:28 quantum computer so in quantum computer
3:30 you have an exponential
3:33 Advantage for typical Quantum sensor I
3:36 mean there are many flavoring but the
3:39 most typical um scaling that we have is
3:43 that if you use quantum s systems as a
3:45 classical sensor or we just have n
3:47 classical sensor the signal to noise
3:49 ratio improves as a square root of the
3:52 number of sensor themselves however if
3:55 we use fanton sensor by exploiting there
3:56 a most fundamental property in
3:59 particular entanglement we can have a
4:01 skin in with the number of sensor which
4:04 is proportional to n and so this gives
4:07 this if you want polinomial improvement
4:10 with respect to classical systems uh
4:12 which is what gives the quantum
4:15 Advantage however reaching this Quantum
4:18 Advantage is uh relatively uh difficult
4:21 and so um when we were working actually
4:24 together with fredman and also Christian
4:27 Gan in trying to analyze and and provide
4:29 a review of what we mean by Quantum
4:32 sensing we came up with if you want
4:35 three classes of quantum sensors uh so
4:37 the first one is where we just use a
4:41 Quantum object uh uh to sense thing so
4:43 if we considering since we were just
4:47 discussing few minutes ago uh MRI of
4:49 course we're using spins in order to
4:51 detect for example the anatomy of brain
4:54 and Spins are intrinsically a Quantum
4:56 object but we can describe them really
4:58 as classical magnetic dipole so we don't
5:02 really use the all uh uh if you want
5:04 quantum mechanics to describe a very
5:07 Basics MRI experiment so we are using
5:09 Quantum object but not really exploiting
5:11 their Quantum property and this is a
5:14 sort of if you want a Quantum 1.0 type
5:17 of technology which we're not really uh
5:19 so interested in defining at least not
5:22 not today for me so what we really
5:24 interested or what we really mean when
5:28 we Define a Quantum sensor are cases in
5:30 which we are using for example at least
5:32 some Quantum property for example
5:35 coherence superposition in order to have
5:37 uh better
5:40 sensitivity and uh even more what we
5:42 would like to do as I said is to use
5:45 entanglement in order to achieve this
5:47 Quantum advantage and of course there
5:50 has been already some uh demonstration
5:53 of this type of sensors so we heard a
5:56 lot about V EnV Center so if you use a
5:58 single MV or if you use an ensemble but
6:00 without using if you want correlation
6:04 among the spins this will fall into this
6:06 category another good example are atomic
6:09 clock so atomic clock are sort of like
6:10 in between sometime people complain to
6:12 me that they should be in the force
6:14 category uh but really we had a lot of
6:17 advances for example brought by Juni and
6:20 an Maria re in which you're using
6:23 Quantum logic to improve your ions at
6:26 atomic clocks you're using Ensemble and
6:27 correlation to make them even better so
6:29 it's sort of like bridging that Gap to
6:33 achieve these really exploiting Quantum
6:36 properties and uh even ligo in its first
6:39 ination was if you want in the second
6:43 category like just interferometer but
6:45 more recently they've been using a
6:47 squeeze light injecting squeeze light in
6:49 order to improve the detection of
6:52 gravitational W so we already have these
6:55 um examples of achieving Quantum
6:58 Advantage with this type of systems and
7:00 really this is the type things that we
7:02 are also trying to do uh to improve
7:04 Quantum sensing really to bring the
7:08 advantages of quantum sensing uh to real
7:11 applications um same as fredman my
7:13 favorite Quantum sensor is the EnV
7:16 Center in diamond and I think that like
7:17 after restock I don't need to really
7:20 motivate why uh this system is awesome
7:23 uh it can really sort of bridge a lot of
7:26 different um modality of sensing and
7:28 different applications so uh looking for
7:31 example at magnetism and the Nanos scale
7:34 looking at effects from Super conductors
7:37 effect from ferromagnetic materials uh
7:39 but with a spal resolution which are
7:42 really down to the nanoc scale um for
7:45 example through Vis scanning tip it can
7:48 look at biological effects uh for
7:51 example some work done uh to detect uh
7:56 HIV RNA uh in um like a a bios uh type
7:59 of setup and also you can look at just
8:01 the Dynamics of other Quantum system for
8:05 example correlation of quantum spin
8:07 Associated either with the diamond or on
8:11 the surface or the Diel so it's really a
8:13 very powerful sensor with many different
8:16 applications and uh well my manx slide
8:20 probably it's a bit superflu since we
8:22 already had an introduction of NV Center
8:25 much better than mine and by now I keep
8:27 it really down to just one slide uh
8:29 again these are just optically active
8:32 defect uh so we will illuminate with
8:35 green light and we see uh the red light
8:38 emitted and uh to put in a plug for
8:40 tlaps which was the two talks ago they
8:42 actually s sell a Quantum diamonds with
8:45 Envy defect so you can also purchase
8:48 that and start doing uh your own
8:51 experiment and of course uh we combine
8:54 typically with Optical uh polarization
8:58 so if you want cool down by Optical
9:00 means and um I used the conventional
9:03 Optical rout for MV centers and we
9:06 combine that with
9:09 microwave control uh where of course
9:11 your Envy Center acts a little bit like
9:15 the compass needle uh pointing up or
9:17 down in response to an external magnetic
9:21 field and you can have magnetic resonant
9:23 driving OD MV and where you detect a lot
9:27 of light if it's pointing up and uh less
9:28 light when it's pointing down and you
9:31 can drop this very nice coherent
9:34 superposition and this is work going
9:38 back to a 2006 B children but people
9:41 have been uh observing uh even better
9:45 coherence time since that time so how
9:48 can we achieve this type of quantum
9:51 Advantage with this type of defects one
9:55 way uh that indeed we propose even back
9:58 then uh in some of the first few
10:01 proposal is to use large ensembles of NV
10:03 centers and trying to exploit it
10:05 interaction among them however the force
10:08 me that appears when you're having large
10:10 ensembles with very very dense and
10:13 Venter is that the interaction among the
10:16 spin V cells uh sort of make your
10:19 coherence time decrease and so you have
10:22 less time to do the detection and so the
10:25 sensitivity so if you want the minimum
10:28 field that you can detect uh is not so
10:30 good and you cannot go for very long
10:32 sense in time because the coherence
10:34 kicks in and you lose your signal so
10:37 what you can try to do is to try to uh
10:39 if you wanted to couple or cancel out
10:43 this interaction and uh some uh very uh
10:44 clever people in Misha looking group
10:48 they uh devise very powerful sequences
10:50 which were able to improve dramatically
10:52 how well you can detect your magnetic
10:56 field with very high uh dense system
10:59 however this does not give you yet um if
11:02 you want um the quantum Advantage per se
11:06 so uh if you were able to change the
11:09 number of mvs uh inside the sample uh
11:11 which uh like in this experiment we were
11:14 not doing uh you would see that the
11:16 sensitivity would just improve as the
11:18 one with the square root of the number
11:20 of sensor at
11:23 most however if we could indeed exploit
11:26 instead interaction among the spin and
11:29 achieve a type of correlation for
11:31 example in the form of a a squeeze state
11:34 of the spin then the sensitivity could
11:36 improve even
11:40 more um we proposed that back in 2009
11:43 and we have not really uh been able to
11:45 demonstrate an advantage we have been
11:48 able to demonstrate a small Advantage
11:50 based on some small entanglement but
11:53 this powerful scaling with this uh which
11:56 improves over the classical level as a
11:58 square root of n has not been able to
12:02 demonstrate yet so um of course uh I've
12:04 been trying to do something else uh in
12:06 between and the idea is that well maybe
12:09 there is something in between these two
12:12 uh limits so we might not just use um
12:15 coherence of single uh Spin and we might
12:19 not look for the full entanglement uh
12:22 enance Metrology but there is maybe some
12:24 regime in between The Chew where we use
12:27 some small um Quantum system uh some
12:29 small amount of entanglement to achieve
12:32 some task that a single Cubit sensor a
12:34 single Quantum sensor could not achieve
12:37 and this is like a very fruitful regime
12:40 which sort of uh is the intersection
12:44 where engineering um advances come
12:46 together with the fundamental quantum
12:49 mechanical advances to bring um some
12:51 real practical application and this is
12:55 what I want to talk in in the rest of my
12:57 talk and there are many different things
12:59 that we have looked uh over the years
13:03 for example one can use some additional
13:05 um Cubit sensor to do Quantum eror
13:08 correction or to use VMS Quantum memory
13:11 in order to both improve the sensing
13:13 time and so the coherence time or to
13:15 look at correlation what I'm going to
13:18 focus today is how to use some
13:20 additional uh control and some
13:24 additional uh level other some virtual
13:26 degrees of freedom or some real degrees
13:27 of freedom associated with other spin in
13:30 the system in order to achieve frequency
13:32 up conversion in order to achieve Vector
13:35 sensing and also to enhance
13:37 coherency so the first thing that I want
13:41 to talk is this vectorial sensing uh
13:44 which is achieved uh by doing some modulated
13:45 modulated
13:49 driving so um before that I want just to
13:52 step uh back and sort of describe some
13:54 very simple way in which you could do uh
13:58 vectorial um sensing so first of all you
14:01 could do the typical Ramsey magnetometry
14:03 to detect field which are in the Z
14:05 direction if you want aligned with the
14:08 NV axis with the MV quantization
14:11 axis the idea here is to prepare
14:14 superposition state of your MV Center
14:16 let it evolve for some time under an
14:18 external magnetic field which is aligned
14:21 in the Z Direction and then detect out
14:22 the phase which has been
14:25 accumulated uh and this will tell you
14:28 how strong of a magnetic field uh it was there
14:29 there
14:31 if you want to instead detect a
14:35 transverse field if you do this you get
14:37 some perturbation but it's very small so
14:39 it's very insensitive way of detecting
14:41 transverse field so what you can instead
14:43 do is to detect field which are rapidly
14:46 oscillating at the resonance frequency
14:49 of your uh spin of your MV Center and
14:52 this will uh indeed induce a a pression
14:57 uh of the spin now in the um Z let's say
15:00 zy plane but this can only detect things
15:02 which are uh oscillating at the
15:04 resonance frequency or the MV Center
15:07 itself so these are gigahertz um for uh
15:10 the Z field you could uh indeed also
15:13 measure oscillating field okay for
15:15 example by using dynamical decoupling
15:17 sequences but the typical frequency here
15:20 on the order of megaherz while here on
15:22 the order of gigahertz so um it's
15:24 difficult to bridge between uh these two
15:27 uh modalities and also the experiments
15:28 are quite different so if you have
15:31 system I if you have experimental error
15:32 they will be different in the two
15:34 experiments and so it's a bit difficult
15:37 to really get out some very nicely
15:42 consistent way of of detecting SP so how
15:44 can we uh Bridge uh these two and find
15:47 the single protocol to do indeed Vector
15:50 magnetometry so before I explain that I
15:51 just want to take a little bit in aide
15:53 so if you do want to do this rabi
15:55 magnetometry how can we understand that
15:57 well we could have this oscillating
15:59 field as I said like a transverse field
16:01 that we might want to measure or use to
16:04 control the spin if you have this
16:06 oscillating field you can always
16:07 consider it as the Su or two counterpropagating
16:09 counterpropagating
16:12 field and then if you jump into the
16:15 frame rotating with the red uh field uh
16:17 this will become of course static in
16:19 that rotating frame and the other field
16:21 will be rotating very fast so we can
16:24 actually ignore it the static field will
16:27 then cause a tipping of your Spin and
16:30 you can detect how fast it is indeed
16:32 rotating and with that you do rabby
16:36 magnetometry indeed uh we we did that
16:38 actually with my colleague Danielle
16:41 Breer of linol lab and achieve some very
16:45 good um picotesla level sensitivity for
16:48 this fast oscillating field uh by using
16:52 their expertise in achieving uh the best
16:54 overall sensitivity in Ensemble like
16:58 large Ensemble magnetometry so this is
17:00 the current sensitivity limit if you
17:05 want for this high frequency field
17:07 detection now going back to our
17:08 oscillating field if the oscillating
17:10 field is along the Z Direction and again
17:13 you jump into a rotating frame in the XY
17:15 Direction well okay nothing really
17:18 happens so it's a it's a bit boring but
17:20 we can use this fact to achieve this
17:23 Vector magnetometry let's say it now I
17:26 have an oscillating field okay which is
17:29 in some um arbitrary Direction in the uh
17:34 in in any direction in Des XY space and
17:37 in the lab frame this is that this
17:40 frequency Omega s but if I look at it in
17:42 the rotating frame rotating at the
17:44 frequency of my apply microwave for
17:46 example then this splits the two
17:48 components so the component which is
17:50 longitudinal will stay at its original
17:52 frequency but the one which is in the
17:54 transverse plane will have a shift of
17:58 frequency by the rotating frame uh uh
18:01 frequency itself Omega and so one can
18:04 sort of detect The Chew in different
18:07 ways and indeed this is what we do we
18:10 detected transverse component uh by uh
18:13 applying for example uh by by looking at
18:15 this Shi the frequency while we detect
18:18 the longitudinal component at the uh
18:21 original frequency and we can indeed see
18:24 them uh in this particular case we see
18:27 that we were applying a field at about
18:30 28 megahertz and this uh can be detected
18:32 the longitudinal component can be
18:34 detected at the original frequency while
18:37 the transverse one is shifted by the NV
18:40 frequency and so we are able to use
18:43 basically the same uh setup uh the same
18:47 NV and the same um experimental um
18:51 sequence to detect two components of the
18:54 magnetic field so the X component and
18:55 the Z
18:58 component the limitation here was that
18:59 as you can see the difference between
19:03 the two is only about 6 mahz and this
19:06 was what the frequency of MV we needed
19:09 to set and that uh if you know the
19:11 technicality that means that you need to
19:13 work at a very specific external
19:17 magnetic field and this is the if you
19:18 want the constraint that you have in
19:20 order to be able to detect both of them
19:23 simultaneously if we make this a field
19:25 frequency with signal frequency much
19:27 much higher then it's not really
19:29 possible to autod detect the transverse
19:31 component with the same type of sequence
19:33 so this is sort of a a bit of a
19:35 limitation but the advantage is that you
19:37 have the same experimental errors and
19:41 you can use a single MV uh to do uh both
19:43 vectorial magnetometry with all
19:46 different axis and so you can also of
19:49 course apply this to an ensemble without
19:51 needing to use different classes of MV
19:52 for different
19:56 direction So to avoid this sort of
19:58 bandwidth limitation that we had in this
20:01 sample uh we decide to use the MV Center
20:03 as a Quantum frequency mixer so what you
20:06 would need is of course sort of like to
20:07 upcon convert or that convert your
20:10 frequency to a frequency which your MV
20:12 is particularly sensitive to and this
20:16 what we wanted indeed to do so again
20:21 like if I have H just an arbitrary um
20:24 field which is oscillating uh I have
20:26 that they are sort of like splitter the
20:28 two components are splited and one is
20:31 has a frequency shift but this frequency
20:32 shift might not be at the frequency
20:35 Target which I like and so what we can
20:39 do is to add some additional bias field
20:41 at an intermediate frequency so that I
20:43 will have an effective signal uh which
20:46 is now nicely situated at a frequency
20:48 typically on the order of a TENS of
20:51 megahertz which my MV can detect very
20:54 very easily with common uh U um
20:57 detection uh um protocols that people
21:00 have sort of like deoy them many many
21:02 times and indeed we do this and now we
21:04 are able to detect for example signal
21:08 which are at around 200 150 megahertz
21:09 which is typically frequency difficult
21:12 for the mvy center but just using a very simple
21:14 simple
21:17 protocols so how does this work well we
21:21 can analyze this by using a FL formalism
21:24 which uh What uh it tells us that really
21:25 what is happening is that we are
21:28 creating some additional virtual degrees
21:31 of Freedom additional uh energy levels
21:34 uh thanks to uh the drive so thanks to
21:36 the drive of our bias field in addition
21:38 to the signal field so we are
21:41 engineering some additional degrees of
21:43 freedom which allows us to do this
21:45 better sensing
21:48 task and the advantage again is that we
21:52 can then use typical um rabi or cpmg
21:54 sensing sequence to detect fields which
21:58 otherwise we will not at the gigahertz
22:01 that we will have needed for Rabbi or
22:04 low meah hortz that we would have needed
22:07 for cpmg the disadvantages that we have
22:10 a heat in the sensitivity uh but of
22:14 course if you cannot sens this sort of
22:17 like 100 of megahertz or or maybe 100 of
22:20 gigahertz frequency in other ways well
22:23 you you can sort of survive um to uh to
22:25 vis a reduction in
22:27 sensitivity so we demonstrated that and
22:29 we demonstrate that we can achieve the
22:32 same level of sensitivity well we we use
22:35 a rabi sensing protocol or a cpmg
22:38 sensing protocol cpng is where you apply
22:41 periodic pip polies which match uh the
22:44 frequency of your target signal and we
22:45 can indeed
22:48 achieve very good sensitivity with this
22:51 produ so this is an example in which we
22:53 just use control to engineer if you want
22:56 this vition additional energy level but
22:58 what we can also do is actually to use
23:01 additional degrees of freedom inside the
23:02 diamond in particular associated with
23:05 the uh nuclear spin in order to uh
23:07 improve the sensing task and actually
23:10 use the electronic spin now just as a
23:13 controller so before I introduce uh the
23:16 Envy Center um and also I think that in
23:18 the freedom introduction we focus on the
23:21 electronic spin but of course we also
23:23 have a nuclear spin in particular the
23:26 nitrogen 14 nuclear spin which is a spin
23:29 one and this has both and hyperfine
23:31 interaction which couples the nuclear
23:33 spin to the NV and also a quadrupolar
23:35 interaction which gives if you want a
23:38 local magnetic field on the the nuclear
23:41 spin itself and we can control also the
23:43 nuclear spin in the same way that we do
23:45 the electronic Spin and again by mapping
23:47 the nucleus
23:50 spin uh um state to the MV Center also
23:54 read it out so this is a much slower RAB
23:57 sorry this is I think rabi illation it's
24:00 better label of the EnV center of the
24:03 nitrogen 14 of the MV Center so we can
24:06 in principle also use this as a a
24:08 magnetic field sensor but it's not very
24:10 sensitive and so what people have been
24:14 using the um nitrogen spin is for uh
24:17 this idea of building a nuclear spin
24:20 gyroscope associated with the NV centers
24:23 because uh nuclear spin are as sensitive
24:25 as electronic spin to rotation and we
24:27 can use the electronic spin to mediate
24:30 the polariz ation without a
24:33 control so uh we uh indeed were looking
24:37 at the nitogen 14 and we wanted to test
24:39 the fact that nuclear spin in principle
24:43 should be uh very very uh uh robust to
24:46 any external perturbation and so I have
24:48 very long coherence time instead we
24:50 found relatively long coherence time
24:53 this is almost millisecond time but also
24:56 some puzzling behavior in particular if
24:58 the MV Center is in the minus one state
25:01 or plus one state the coherence time was
25:04 much shorter on the order of 2
25:06 millisecond instead of almost 1
25:08 millisecond so what we figure out very
25:10 quickly is that actually variation in
25:12 the hyperfine and quadrupolar
25:15 interaction were reducing the coherence
25:19 time so one can sort of use some Eon on
25:22 the Envy Center to correct for
25:24 this but we were also interested in
25:26 trying to find some other system which
25:28 hopefully didn't have this problem at
25:30 all and so what we did was to turn to
25:34 N15 so N15 is a spin 1 half isotope so
25:36 it does not have any quadripolar
25:38 interaction so well it should have a
25:40 better coherence time and there is no
25:42 hyper fine if the NV is in the MS equal
25:45 zero state where this spin is inde zero
25:48 and so what we expected was that as we
25:50 were going from the minus
25:53 one state for the EnV Center and looked
25:55 in at the coherence time of the nuclear
25:58 spin so if we move to the MS equal Z we
26:00 should expect much longer coherence time
26:03 instead it was a bit longer but not as
26:05 much as we were
26:08 expecting so again this was a brief
26:12 puzzle until we of course figure out if
26:14 I have some water I don't have water we
26:16 figure out that actually there is some
26:18 hyper fine because we have some
26:20 transverse component of the hyper fine
26:22 and as soon as you have a little bit of
26:24 external transverse magnetic
26:27 field with this coupled in in a very
26:29 very effective
26:33 way because that I need to find water
26:34 otherwise I'm not going to be able to continue
27:01 sorry about that let's see if it is
27:11 not so what we found was that transverse
27:14 field couple into the
27:17 hyperfine giving rise to a shift in the
27:19 resonance frequency and if the hyper
27:21 fine is in homogeneous then you have
27:24 again some Decay and this shift can be
27:27 quite large especially around uh the
27:30 level The Crossing for the nucleus Spin
27:34 and uh and of the electronic nucleus
27:37 spin so what can we do well again it
27:40 becomes better actually to uh sort of
27:42 alternate between the minus one and plus
27:45 one states of the electronic Spin and
27:48 this will actually improve the coherence
27:51 time of the nuclear spin itself because
27:54 this will sort of refocus the hyperfine
27:57 neogenes that we overwise obsorb and
28:00 indeed if you look again at the nuclear
28:02 spin coherence Time by doing this
28:05 protection scheme we can uh improve the
28:10 coherence Time by almost an order of
28:13 magnitude so the last uh thing that I
28:15 want to say is that we can actually use
28:17 the electronic spin also as a Quantum
28:20 amplifier again I told you that we have
28:23 this announcement of a transverse
28:27 magnetic field uh because of the
28:29 coupling with the hyper fine between the
28:32 nuclear and electronic spin so the
28:35 nuclear spin itself will see a much enan
28:39 uh uh transverse magnetic field and this
28:41 can be used to detect a better a
28:44 rotating magnetic field or if you want
28:46 if the diamond itself is rotating with
28:48 respect to a static magnetic field uh
28:51 the nuclear spin will see a much larger
28:53 rotation than the electronic spin itself
28:56 so what we can do is then to exploit
28:58 this fact to have an anouncement in the
29:01 sensitivity to rotation sensing so as
29:03 you can see the rotation of the diamond
29:06 is very slow but the nuclear spin see a
29:09 a much larger change and uh its Dynamics
29:11 can just be a still like AA following of
29:14 its initial state so it's very robust to
29:17 additional noise but it can give you a
29:20 very sharp advantage in the
29:23 sensitivity so what I show you that we
29:27 can use the electronic spin uh both as a
29:30 Quantum mixer as a a Quantum controller
29:32 and as a Quantum amplifier so all type
29:35 of quantum Advantage which we can get
29:38 from our Quantum sensor we really
29:41 needing to go for very large quantum
29:43 entanglement so this tells us that
29:45 indeed we can already get some Quantum
29:48 advantages although of a slightly
29:50 different time for hopefully near term
29:53 technology so even beyond the isber
29:55 limit even if you have some small amount
29:57 of entanglement or some additional
29:59 degrees of freedom we can make Quantum
30:01 and un sensing much more practical so
30:03 for example we can have a frequency
30:06 conversion of this cerence protection or
30:09 amplification of some external
30:12 perturbation so what are other
30:13 challenges of opportunity for Quantum
30:17 sensor I think that what PR describ in
30:20 the previous uh uh talk that was
30:23 perfectly spot on uh what we want to do
30:26 is really to uh try to bring all these
30:29 devices from our Labs out in the world
30:30 and there is still a lot of work to be
30:33 done there in particular in trying to
30:36 make uh sensor which are much more
30:38 robust so that they can
30:42 work in um the condition that people are
30:44 really interested in and I think the
30:46 other thing which is quite interesting
30:48 nowadays is to try to integrate Quantum
30:50 sensor in the larger Quantum ecosystem
30:53 so really uh integrate Quantum sensor
30:55 into Quantum networks integrate Quantum
30:57 sensor with quantum computer so that we
31:01 can uh really uh both exploit for
31:04 example distributed sensing and also uh
31:06 try to use a small Quantum compitation
31:09 to improve the Sens in task itself with
31:11 this I want to thank uh the people who
31:14 have uh done uh the work in particularly
31:18 minty and Goin Wang who not in this
31:21 picture because he already graduated and
31:22 uh thank you for your attention and
31:39 over one second sorry this is very s you can
31:40 can
31:44 take okay can can you hear
31:46 it uh do we have any questions for Professor
31:56 question we are little we they can uh we
31:58 can answer the there's a some question
32:01 in the Q&A but
32:04 maybe Professor can can answer those
32:08 offline yes because we a bit behind Okay
32:10 and we can start thank you so much thank
32:12 you so much I'll stop my sharing thank
