Embedded Radar Prototyping | S. Bottigliero | PitchD 01

Embedded Radar Prototyping | S. Bottigliero | PitchD 01


thank you so good afternoon i’m Stefano
and so first time of like collision avoidance system this is the work that
follows my master’s thesis and what is a time of flight collision
avoidance system well the basic diagram of the time of flight system is a
composed of a transmitter that in our case would be a laser diode that will
send light to a lens that will diverge the beam laser beam in order to
illuminate a larger area then the environment will scatter back the light
to a lens that will focus the light received on a time of light sensor a 3D
sensor that is de-modulated with the same sensor that is used in the modulation
and the result of this computation is a phase value that we use in order to have
a depth map of the environmental of the scene illuminated by the by the by the
laser so what have I done well I designed this board that is the result
of the integration of a different board we used to evaluate the prototype at the
beginning and the goal was to have a more compact system that merges the two
we had a border for the sensor and the border for the FPGA an Altera FPGA in
order to do some signal processing and filtering on the receiver data phase
data so the idea is we illuminate the environment we received the phase
information from the de-modulated light by the laser and we have a 3d map that
we can filter and on which once we have processes processing it we can apply
some threshold algorithm in order to decide if for example an object now in
the phase we know the relation between the distance of the object and the phase
they received so we can evaluate a it’s distance from the sensor and of
course if we apply a trash chute we can say okay there’s something in this frame
that is too close to the sensor we can apply some counter-measure the
idea is to use the this system this system is thought to be used on
automated guided vehicles that are a so-called the ATVs so in this way we can
have something that is automated that can move freely in space and in real
time it and it can have a map a 3D map of the environment so you can can be
able to take decision for example to emergency braking or steering in certain
to avoid collisions which are the result well the color are not so good I’m sorry
for that but in the central image on the top you can see the phase received by
the sensor and the send to the FPGA and in the lower part you can see the
filtering applied in software and on the top left you can see the same filtering
applied in hardware so after that we have an image that if you look I don’t
know if the colors are well rendered but the red part the the note something that
is closer to the sensor then something that is a farther like in orange or blue
you can see there’s my chair in orange a at the center of the image and we have a
red beam in the front of the scene and that’s it for I would like to underline
I designed both the hardware the the board I showed there and the software
and the manipulation of the face data in a in hardware that concludes the first
project I’ve been working on and I’m still working on the second one is the
development of a real-time locating system the idea comes from the fact that
suppose you are in a warehouse you have any TV and you want to know indoor where
press where these machine is well using GPS
may be tricky sis you you can have problems in
reception so the idea is to use a system that is able to track an active target
that can be attached to the machine and can be located with respect to fix at
the same two sensors that are in fixed position that we know so we can evaluate
the difference of the position sent by the target and that one received by the
sensors the principal we are using is the time of arrival so we send a wideband signal that is received by the sensor by the sensors and since the
sensors as are synchronized we are able to evaluate the time difference of
arrival between each sensor and the master sensor in this way we can drill
at a rate at the position of the target in the environment this is this allows
real-time location using correlation inside again and FPGAs and we can map
the position both in 3d in 2d or in 3d to have the 3d localization we need the
more sensor more at the basis the system is able to have resolution of an
accuracy of 10 centimeter it depends of the sampling frequency we are using what
I’m doing right now is developing the film where part both in the in vhdl for
the on exciting FPGA and the code for the processor since these tilings a7000
we are using is an SOC so we have both FPGAs and the processor inside the same
chip so it is appropriate to for an embedded system since we have everything
on just one chip and I’m also designing the new tag that can be that should be
low weight very low cost and can be easy to manufacture the idea is to develop
out the system on this evaluation board and then again create a new board that
integrates all that you see in this border plus the new part the receiving
part at RF radio frequency all in one system that again should be very compact
and low cost the last part of my discussion is my contribution in PCB
rather this automotive radar design so in these pictures we are collaborating
naturally with the STMicroelectronics in developing both rather at 77 gig earth
and another automotive era there are 24 Giga that are the two bands allocated
for automotive Rider in the pictures you can see on the left a border that I
designed at 77 gig art they are not in scale so I couldn’t say you can
recognize it because the antennae are smaller but you can see that there’s a
lot going on we have many antennas in this case we have 12 receiving antennas
and 9 transmitting antennas and it works a 77 gigahertz the other one is a 24
Giga Hertz board that has for transmitting antennas arrays of antennas
and to receiving one then you can also see that there are some balloons so in
order to modify the senior from differential exiting from the chip to
single ended and it is a I think a good example since it gives it gave me the
opportunity to see both something that is come is related to radiofrequency
design in terms of design in the PCB because I wasn’t I wasn’t capable of
design in the antenna I had a more expert colleague much more expert
colleague than me that design it for for the border
but you have the opportunity to design both a system at a radio frequency that
has an analog part and the digital part and so make a PCB that merges almost
everything you have seen in in the courses during your master degree or
bachelor and I think it’s a they are very beautiful from this is my opinion
last things this concludes what I’m actually doing right now what I’m going
to do the idea is to take the two systems that I shown you the real-time
locating system system and the collision avoidance system and integrates them
into so-called simultaneous locating and mapping system slam in order to have a
navigation system that is able to create a 3d map of a scene for energy V in
order to move freely in in space and also to be able at in real time to
locate the same object with a good accuracy in the order of 10 centimeter
then the idea is to continue also on the automotive radar on PCB development but
not only also try to develop and work with the development of the system of
the femur for the for the rider and the different application of techniques to
benchmark and validate and validate it and that’s it I I have no other words to
show you for now for now and I finished you

Leave a Reply

Your email address will not be published. Required fields are marked *