-
- 30 Nov 93 Kevin Wasserstein, Hughes Liaison
- Did not know reason for 45 degree down tilt.
- Informed him that we want to visit next quarter, winter.
- Notified us that aperture is a window
- Received Fax of his pitch slides
-
- 30 Nov 93 Alex Kormos, Head of Engineering for NIGHTSIGHT Project
- He will fax or fedex specs and drawings by COB 3 Dec 93.
- The reason for the 45 degree down tilt is marine applications
- Field of vision: 28 degrees horizontal, 18 degrees vertical
- 360+ degree pan is a good idea, look at cost.
- Aperture or window is 4" by 2", all light goes through it. Thermal radiation works the same way as visible light.
- 11 Jan 94 Victor Sheinmann
- Yaw may not be a desirable degree of freedom. Because it does not increase the field of view, yaw may be easier to produce through electronics or software.
- His philosophy is that problems should be solved with the following tools; software, electronics, and mechanics in that order.
- Thought the power requirement was not too restrictive as long as the pan and tilt mechanism operated on the center of gravity of the sight. In this way the power only worked against inertia and not gravity.
- 31 Jan 94 Kevin Wasserstein
- Kevin gave the following motor companies as possibilities:
- Micro Mo -miniature precision motors (high cost)
- Globe Motors (513) 228-3171 - Little expensive, but will to work with you
- Auto Troll (sp) - Cheap (highly recommended, but no phone number
- Johnson Electric (203) 255-5761
- Pittman (215) 256-6601 (In Pennsylvania)
- Naguchi (sp) - really cheap, but maybe too small
- Douglas (708) 232-7419 - cheap
- I also mentioned to him that for the police car application, the hand crank may be the most economical way to go. I asked him what percentage the pan and tilt mechanism is of Hughes\q cost for the night sight package. He said Pan and Tilt might be about 1/3 of the total cost to Hughes, with a mark-up of 3-4x the cost by the time it gets to a law enforcement agency. If it\qs really one third of the cost then we may want to have that be a second option in our final spec.
- Kevin cautioned us to make sure we order parts from venders soon enough of them to get here. Steve updated Kevin on the current design track we are pursuing of keeping both motors on the non-moving base.
- We agreed to call him again next monday at 1:30.
- 4 Feb 94 David Downs, Vice President of The Motion Group, Inc.
- The Motion Group manufactures control devices for mechanisms. They specialize in software controlled motors that can give very specific input to the motors in terms of position, speed, and acceleration.
- We discussed the possibility of different types of interface. David prefers a track ball. If we want to use a joystick, their are two options:
- The joystick controls pan by moving the stick left and right, and tilt by moving the stick up and down.
- The joystick only controls pan by pointing the stick in the direction you wish to look. A separate mechanism is needed, like a thumb switch, that will provide for tilt.
- The control boxes range in price based on the number of bits you desire. For the base line box, 4 bits, the cost is $450. We would require two boxes for the Type 2 Joystick, one for each degree of freedom, unless we were willing to give up a bit.
- The advantages are that this company will handle all of our control needs and the sight would easily meet the requirements we need to meet. Also, this company promises to be a good source of stepper motors and because, of their location, we'll be able to get them quickly.
- Regarding controls, they are not to be included in the $300 price specification. Motors are included. Kevin stressed, however, that the package cost should still be minimized. Alluded that $400 for a complete system with smarts would not be considered a bad solution.
- Mentioned that he would try to get more information regarding the design their working on at Hughes.
-
- 24 Feb 94 Stanford Sheriff Department, 723-9633 / 329-2413 (night)
- In order to determine the needs of potential users, we requested a night time ride-along with the Stanford Sheriff Department. Steve and Joe rode with two separate officers for approximately 45 minutes while Jim spoke with the Desk Sergeant.
- SGT Love, Desk Sergeant:
- Interface: SGT. Love thought that it was easier to manipulate a lever device than a knob, he indicated that a joystick would be preferable. He also thought that the device should be mounted below the dashboard level so that it could be used discreetly.
- Field of Regard: SGT. Love did not see a great need for greater than 340 degrees of pan. He indicated that police always try to keep their vehicle oriented in the direction of action, and that if he were not pointed that way, he would move the car.
- Robust: Indicated that the device would need to be very robust. He mentioned that vehicles had been vandalized in inner cities and in riot or crowd control situations. Also, the cars are subjected to very difficult weather and road conditions.
- Detachable: Indicated that it would be great if the device detached easily from the automobile. This would make it easier to secure and account for the device. He also indicated that it would be beneficial to be able to use the sight in a hand-held mode, but that\qs Hughes\q problem. With the concept of community policing, officers spend much more time on foot.
- Maintenance: Agreed that the device would have to be almost maintenance free. Police don\qt do a great deal of their own maintenance. For the rest of their technical equipment, they have civilians that perform the user level maintenance, and they contract out for the higher level maintenance.
- Other Contacts: SGT. Love gave us several further contacts:
- Coast Guard: Maritime operations
- East Bay Regional Parks: Maritime Operations
- Palo Alto Fire Department: Maritime Operations
- Tom Sing: Assistant Sheriff, Santa Clara
- LT. Brunot: Stanford Sheriff Department
- Overall Impressions: Some may argue that a police department as small and quiet as the Stanford Sheriff Dept. would not be a great place to get feedback from the true potential users of our product. However we found the police to be professional and extremely knowledgeable. They were interested in the potential of the sight and had definite opinions on how it should be used. Overall, the ride-along was an extremely fulfilling event.
- Officer Lavoy:
- His background includes working in several police departments. The Stanford police vehicle seemed to be relatively well equipped (lap top computer, three search lights, expensive police radio, Unitrol light and siren controller).
- We performed the regular patrolling duties but focused on patrol and surveillance techniques. Some of the important points discussed and lessons learned are the issue of left-handed versus right-handed users. The current design of the center mounted spotlight requires the user to operate the device with his/her right hand
- The tactile feedback of the top mounted lights is very good but, the lights generally leak. It is common that the officers hang on to this light handle when going around corner at high speeds and since it is only mounted to the sheet metal roof, it tens to become loose and leak.
- The dashboard and center console area seemed very cluttered. There is definitely a market to make an interface which is much more friendly.
- Younger officers would probably be comfortable with a joystick controller design given that many of them grew up with video games. The older users would probably prefer the tactile feedback of a mechanical linkage more like the current design.
- Be wary of a steering wheel mounted controller. With current safety regulations and the positioning of airbags, keeping the steering wheel clear of controller and gadget might be a good idea.
- Device might be used to help police officers in backing up situations.
- During a surveillance operation, it is generally the preferred technique to back into a position. Therefore, a device which scans unlimitedly around was not considered necessary. It occurred to me that a scanning mode whereby the camera could be scanned around automatically so the operator wouldn't have to move a controller back and forth.
- One of the best aspects of the current design is the fact that the panning position of the spot light can be predicted before the light is turned on. The face of the palm corresponds to the position of the head.
- An easily detachable and transferable system would be very convenient since the turnover of police vehicles is only about 12-18 months, and it is not uncommon to have a disabled vehicle. It would be a waste to have a $1200 system that cannot be used because it is in a disable vehicle.
- A system which utilized the current computer monitors would save space in the very cluttered environment of the patrol vehicle.
- Tactile feedback is great, but cops love gadgets.
- Officer (?) [Woman, Name Possibly begin with A]:
- Suggest going to a "cop shop" like Yorkshire to look at all gadgets that police buy and how much more they are willing to pay for new gadgets.
- She said that though manual control might give immediate tactile feedback but police would prefer motorized gadget and would be willing to pay 30% more for that e.g. savings on manual wouldn't be worth it.
- She stressed how many police fail to look behind them or up when they are scan an area. This contrast with what Steve was told by the older officer that police rarely look behind them.
- Mentioned that at night, she often patrols on foot because spotlight doesn't go very far and warns of her approach. When it's foggy at night, they can often hear bikes being rattled and then the big question is which bike rack is being hit.
- Probably wouldn't be able to continually look at the monitor when driving so a high contrast button on the sight which would show a person but not much background might be nice so that you could notice something on the monitor from the corner of her eye.
- Criminals always run away from the police car so there is rarely a reason to quickly pan to track someone.
- If parked, it would be nice to be able to pan around to see what's happening without making a give away noise.
- Generally you don't look up so it wouldn't be that much of a disadvantage to have the tilt be in a separate more awkward position with panning very easy.
- A joy-stick left-right (pan) up-down (tilt) interface would mean that you'd have to watch where you are as you turn--thus what you want is a control that says this is the position you are. Thus, imitating the control handle of a spot light seems an optimum design.
- Few police got out with partners
- Mix of responding to calls versus patrolling around. She said she preferred going around on foot because can't find trouble in the car.
- Ideas:
- Control is lamp like and dial to desired position and camera catches up or handle always is with camera angle
- Control sturdy enough to be used a brace for turns.
- Handle is perpendicular to view: Thus when wrist is turned for tilt, the motion corresponds to camera motion. Disadvantage is that its easier to pan to the left than the right, or visa versa.
- Make light part of Nightsight so that when you turn on light is precisely lights up where you were looking.
- Make Pan and Tilt easily removable.
- 25 Feb 94 Norman Stiepel, Sensormatic Electronics (305) 420-2483,
- Norman Stiepel is the designer of a pan & tilt mechanism for surveillance cameras that have many or the same requirements as our project.
- Total manufacture cost of their surveillance camera is $1800. The following is a breakdown:
- Motors (Micro MO 22 series: $300
- Gold Plated Slip Ring: $100
- Camera: $450
- Lens: $250
- He used the Micro Mo 22 series motors because they have and iron-less core. They had a low rotational inertia and low motor noise. Low motor noise made the controllers even more accurate.
- Used HC11 Motorola chips to control the pan and tilt motors. This chip controls the acceleration and positional accuracy very well.
- During testing, he used photonic sensors to measure how much his design resonated. We should probably talk to him again as we get into the testing phase.
- His design pans at 240 degrees /second.
- His biggest hurdle was designing everything so that he maintained a well-defined center of balance.
- Asked about life of the Micro Mo motors. He indicated that they had a very long life (25 million camera rotations) and he thought the bearings or gears would wear first. He recommended large shaft diameters because that would allow for larger bearings and less wear on the gears.
Another of the design options under consideration for the Pan and Tilt mechanism is the U-Joint mechanism, similar to the designs commonly used to move lights at rock concerts. This design consists of two horizontal rotational joints, off-set by 90 degrees from one another. There are two chief advantages to this design:
- It is a relatively straight forward designs with the potential for a nearly off-the-shelf solution.
- The cord of the NIGHTSIGHT camera never gets twisted, allowing unlimited panning of the camera.
FIGURE 24. U-Joint Design
The principle drawbacks of this design are:
- The entire mass of the NIGHTSIGHT camera, in addition to the mass of the Pan and Tilt mechanism must be moved.
- Since the camera never undergoes rotation, the image received by the camera when panning requires correction for the tilting of the image.
The two mirror design was a product of the desire to achieve 360 degree scanning ability without rotating the NIGHTSIGHT camera itself. This, we hoped, would allow for unlimited panning (rotational) capability while eliminating the possibility of wound-up cords or the need for expensive slip rings. The concept was not unlike that of a periscope where the first mirror inverts the image and directs it downward and the second mirror reverts the image and projects it towards the eye of the viewer.
FIGURE 25. Periscope Design
Notice in the design above, however, that the user rotates on a platform so as not to undergo any motion relative to the two mirrors. Also note in the oversimplified picture above that the surface of the two mirrors are parallel.
The two mirror design seen below challenged both of these principles by allowing the camera to remain fixed while rotating the upper mirror relative to the lower mirror to provide panning capability. Tilting was achieved by moving the upper mirror relative to the rotation ring via the two vertical lead screws.
FIGURE 26. Two Mirror Design
Although this design allowed for very little rotating mass (the upper mirror), it presented a serious problem regarding image distortion. That is, as the upper mirror was rotated so that it was no longer parallel to the lower mirror, the image seen by the camera would undergo a rotation. This could be compensated for by electronically altering the image before it is transmitted to the monitor or perhaps by rotating the camera simultaneous to rotating the upper mirror. In either case, we realized through rapid prototyping that our "key" solution would require some fancy electronics or at least, rotating the camera.
A solution for the problematic rotation of the images found with the mirrors designs is to use a dove mirro or prism (see Figure 27).
FIGURE 27. Dove Mirror
TABLE 20. Dove Mirror Advantages
---------------------------------------------------------
Advantage Reason
---------------------------------------------------------
Simplicity One piece solution to image rotation problem
Cost No electronics involved
---------------------------------------------------------
TABLE 21. Dove Mirror Disadvantages
----------------------------------------------------------------------
Disadvantage Reason
----------------------------------------------------------------------
Space Final mirror or lens size may be large
Requires optics Size of mirror/lense directly related
redesign to distance from focal point
----------------------------------------------------------------------
As a possible solution to the problems with the two mirror design, the one mirror design was developed. The design consists of the sight on a rotating base and one mirror above it that rotates with the sight but can tilt independently. Because the mirror rotates with the sight, the "rolling image" was eliminated. The use of one mirror cause the image to be reversed so that left is perceived as right and vice versa. This problem can be overcome electronically by reversing the columns on the display so that the user notices no mirror effect.
The pan requirement will be met by the rotating base. This motion will be powered by stepper motors or servo motors. Many options are still under consideration. If the design is to pan more than 360 degrees , a slip ring or a similar mechanism will be required to accommodate the electronic signal.
The mirror will fulfill the tilt requirement. Once again stepper motors may be used to give the mirror the necessary motion. The use of the mirror is desirable because it will require less power and smaller motors to tilt.
The entire system would be enclosed in a housing that would protect the mirror and exposed electronics. The housing would reduce the amount of maintenance and cleaning the mechanism would require.
FIGURE 28. One Mirror Design
In mid February, we met with Professor Larry Leifer in order to ask "Management" questions about our project. The following are the questions we asked and Prof. Leifer's responses.
-
- We still have not made meaningful contact with any "users". It appears that Hughes has not either. They do not want us talking to local police. What can we do to receive meaningful user input?
- It is imperative that you talk to users. Do whatever it takes to get feedback from the people that will be using your product.
-
- The NightSight has several applications; police car, marine, security. Each application would ideally use a different design. Our current plan is to design for a worst case. Is that the best course of action?
- Prioritize which user to design for and insure that Hughes agrees. Some components will be universal. Get the gimbal system that police departments are currently using
-
- We think the requirements cannot be achieved for the target price. If this is true, we plan to make cost our driving objective. Is this the right approach?
-
- From our plant visit we found out that many of the "requirements" are really goals.
- This is not unusual and is a typical experience in the real world.
-
- A current alternative is to try to use one motor instead of two to pan and tilt. Given that this will make the mechanism more complicated, is this a worthwhile goal?
- Recommended that we look into remote servo motors like the ones used for model airplanes.
-
- We have decided that our design will not cause Hughes to alter their Night Sight design at all. This approach has caused us to abandon the mirror idea and almost restricts us to gimbal type mechanism. Do you agree with this approach?
- Fine, there is still a great deal of engineering involved in designing a gimbal mechanism.
-
- What are Stewart platforms?
- Showed us a magazine article about Stewart Platforms. Told us to contact Mark Bolas, a designer of pan and tilt mechanisms.
The following is a list of vendors who have been contacted as of March 14, 1994.
United Motion Technology --- (408) 737-7001
Servo Systems Co. --- (800) 922-1103
The Motion Group --- (415) 969-5829
Kollmorgen Motion Technologies --- (714) 581-3626
Micro Mo Miniature Drive Systems --- (813) 822-2529
DC Micro Motors --- (708) 232-7419
BEI Motion Systems --- (619) 744-5671
Johnson Electric North America --- (203) 255-5671
Elcom Pittman Brushless DC Servomotors --- (215) 256-6601
Nippon Pulse Motor Co., Ltd --- (703) 639-2495
Barber-Colman Company --- (815) 397-7400
Canon Precision Inc. --- (516) 488-6700
Philadelphia Gear Corp. --- (215) 265-3000
Chicago Gear Works --- (800) 343-3652
Nordex Incorporated --- (800) 243-0986
Stock Drive Products --- (51^) 328-0200
W.M. Berg Inc. --- (516) 596-1700
Secs Inc. --- (800) 533-SECS
Boston Gear --- (617) 328-3300
Martin --- (213) 685-3150
Albro Gear & Instrument --- (800) 832-4229
Bayside Precision Gearheads --- (516) 484-5353
Litton Poly-Scientific --- (800) 336-2112
Superior Carbon Products --- (601) 253-2442
United Equipment Accessories, Inc. --- (319) 352-2175
Michigan Scientific Corporation --- (616) 547-5511
SKF Industries, Inc. --- (415) 943-1800
The following is a list of technical have been accessed as of March 14, 1994.
- The Gyroscope, James B. Scarborough, Aux 531.34.S285
- Problems of Rotary Motion, John Gross Barnard, Green 506.S664 v.19
- The Gyroscope, Frederick Cordeiro, Aux 531.34.C794
- Klystron Reports and Patents, Spec Coll - Univ Arch SC 372
- Special Collection, Green Library , M-F 9-5
- An Elementary Treatment of the Theory of Gyroscopic Motion, Green QA862.T7C8
- Gyroscope, James Reaney, Green PR9199.3.R4G97
- Klystron Technical Manual, Green QC544.V3S6
- Analysis of Rotary Motion, John Gross Barnard, Aux 531.34.B259
- Engineering Design Methods, Nigel Cross
- Design Methods, John Chris Jones
NIGHTSIGHT Pan and Tilt
10.0 Schedule