Description

Bicyclists typically like to follow the smoothest of roads and paths. Small bumps of any kind are jarring and slow you down.

• How much to bumps in roads of different types slow you down?
• How can we identify routes with lots of bumps?
• How can we incorporate this information into route planning?
• How can we use this information to reduce the number of bumps on bicycle routes?

Prior Art

Proposed Methods

Required Resources

Description

I've crossed Stockton at Sherman Way on the order of 750 times. My guess would be that at least 75% of signal activation at least one motorist violates their expected behavior given the signal lights. This is mainly running the blinking and solid red lights, but also includes not going when light stops, stopping at the blank light, turning into Stockon from Sherman Way during the signal, etc. A conservative estimate would be that 50% of signal activations auto drivers do not do what they are supposed to.

My hypothesis is that motorist compliance at the beacon is far less than compliance at a standard red-yellow-green signal or a stop sign or flashing red lights. Additionally, I think the beacon confuses the drivers due to its poor design. Some of the issues are found here:

https://en.wikipedia.org/wiki/HAWK_beacon#Conflicting_meanings_of_HAWK_signal_aspects

Prior Art

There are studies of the HAWK beacon. They generally show that motorist compliance in stopping at a crosswalk for a pedestrian is much higher for a painted crosswalk with a beacon than without the beacon. I have yet to find someone comparing the compliance (over time) so a beacon and to other well known devices: stop sign, red-yellow-green signal, flashing red lights.

Proposed Methods

Required Resources

Description

There are a number of trajectory optimization packages out there now. It would be cool to compare them all for speed and accuracy for solving the basic minimal energy walking problem.

Prior Art

Proposed Methods

Required Resources

Description

Dockless systems require lots of maintenance, from battery recharging to repair to relocation. All of these require people to ride around town (often in motor vehicles) to address this.

Prior Art

Proposed Methods

Required Resources

Description

It would be quite nice to have a place for researchers to upload time series data from experiments with single track vehicles.

Prior Art

Proposed Methods

Required Resources

Description

The fire department and medical resource often demand that roads accomdate their large vehicles at maximum travel speeds. This means that these vehicles fly through neighborhoods and the streets are then designed in a way that other vehicle go faster too. These departments oppose speed bumps/humps/tables, traffic circles, and traffic calming. In the US the fire trucks are especially enormous so the streets have to be enormous. I'd like to put some objective numbers on the consequences of slowing down and making these vehicles smaller. What are the actual consequences to the human lives that are being rescued?

Prior Art

Proposed Methods

Required Resources

Description

Large trucks (including SUVs) are marketed towards non-commercial customers in some countries (esp. the US). These vehicles have become more numerous over the recent years and a growth in size of average vehicles has been well documented. These vehicles are designed in a way that increases the likelihood of killing and maiming vulnerable road users: pedestrians, bicyclists, and even those who drive in small cars. These large trucks have proportionally larger kinetic energy to dissipate in a collision (vastly dominating that of pedestrians and cyclists), the vehicles can drive over people with little notice to the drivers, and the large, high vehicle creates large blindspots that are detrimental to pedestrians even of adult height. There are both scientific articles and pop media sources highlighting this issue over the last decade or so, yet the problem seems to grow.

These vehicle are designed in a way that is detrimental to public health. Some things that we could probably contribute to this conversation:

• Design guidelines for vehicles to minimize risk to vulnerable road users. We should be able to quantify the risk to some degree based on blindspot locations and size, vehicle geometry, vehicle protrusions, ability to squish people, torque/power, etc.
• Vulnerable user danger ratings of existing vehicles based on some metric that quantifies the various.
• Make use of crash data to tie injury to specific design aspects of the vehicles involved.
• Evaluate different designed elements of the vehicle for danger.
• Design vehicles optimal for minimizing risk to vulnerable road users (given some reasonable constraints, i.e. vehicle can't just be a slow moving soft bubble).
• Evaluation of crash avoidance systems and their relative efficacy to simply reducing vehicle size and danger.
• Costs to said drivers of large vehicles if they are banned. What are the negatives of restricting these vehicle types?

Prior Art

An extensive review of the existing work on this topic might reveal some new areas to focus research on, as it seems there is work related to the ideas written above.

• "On foot, at risk" Insurance Institute for Highway Safety: https://assets.documentcloud.org/documents/4453716/Insurance-Institute-for-Highway-Safety.pdf
• https://usa.streetsblog.org/2018/05/09/study-links-rise-of-suvs-to-the-pedestrian-safety-crisis/
• "Effect of Vehicle Front End Profiles Leading to Pedestrian Secondary Head Impact to Ground" https://doi.org/10.4271/2013-22-0005
• " Pedestrian Injury Outcome as a Function of Vehicle Market Group in Victoria, Australia" https://doi.org/10.1080/15389588.2014.1003819
• "Effects of Vehicle Impact Velocity, Vehicle Front-End Shapes on Pedestrian Injury Risk" https://doi.org/10.1080/15389588.2012.661111
• "Effects of Vehicle Size on Pedestrian Injury Pattern and Severity: Prospective Study" https://doi.org/10.1007/s00268-012-1797-4
• "Impact speed and a pedestrian's risk of severe injury or death" https://doi.org/10.1016/j.aap.2012.07.022
• "Pedestrian Deaths and Large Vehicles" https://www.justintyndall.com/uploads/2/8/5/5/28559839/tyndall_pedestrian.pdf

Proposed Methods

Required Resources

Description

It is clear that if you only have steering input to a bicycle that you must countersteer* to change direction, i.e. steer first left, then right to go right. If you ride without hands and let the steer be free you can still change the direction the bicycle travels with a variety of "body language". For example you might say you can lean left or right to change the travel direction. Or you can shift your knees left and right to affect the direction of travel. There are a couple of questions that don't really have definitive answers regarding non-steer control inputs.

1. Can you change the direction of travel while riding no-hands without countersteering occuring?
2. Can you change the direction of travel while riding hands-on without countersteering occuring?

One piece of evidence I have that points to the answer of #1 is this figure:

from Hess, Ronald, Jason K. Moore, and Mont Hubbard. 2012. “Modeling the Manually Controlled Bicycle.” IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans 42 (3): 545–57. https://doi.org/10.1109/TSMCA.2011.2164244.

If you take the Whipple-Carvallo model and control it with a roll torque only (torque between rear frame and ground about roll axis), it seems to require an ever so slight countersteer. A roll torque isn't exactly the torque one would generate between the rider's torso and rear frame, but I think it is effectively the same.

To answer the question we'd need to demonstrate changing direction without countersteering (or find someone who has).

*there is at least one other definition of countersteering: in a steady turn you may have to apply a steer torque in the opposite direction of the steering to maintain the steady turn.

Prior Art

Proposed Methods

Required Resources

Description

Prior Art

Proposed Methods

Required Resources

Description

Prior Art

Proposed Methods

Required Resources

One of my biggest complaints about open source software is the fact that APIs do not remain stable. If I create a research paper using a software stack, publish, don't maintain it, and then come back ~1 year later it seems to take a day or more to update the software such that it can function with the updated dependencies. One year isn't that long of a time in a research world. This isn't good for reproducibility and I don't think we should have to shop a VM with a paper that freezes the entire stack. I've also noticed that my Matlab code that is 10+ years old tends to run just fine on new version, leading me to believe that Mathworks takes this much more seriously.

I'm interested in characterizing:

• how quickly changes in downstream dependencies break scientific software
• the ranking of stability in API for core software packages
• comparing the API stability culture among languages, e.g. Python and R
• how deep in the stack do you have to go to get stable APIs (for example the Linux kernel API is probably rock solid stable)

Hypothesis: On average a given script or software package that relies on a high level scientific computing software stack will break within a year due to unstable dependency APIs.

Prior art

Haven't found anything much yet.

Methods

Here is an idea for a method to do this:

1. Download a package or script at the top of (or near top of) the stack and log its release date
2. Install the dependencies specified at the time of release and ensure the software runs
3. Increment the dependency versions in chronological order and test if the script/package still runs at every increment. You can detect whether is runs or not and also whether deprecation warnings are emitted. If a single dependency fails, you can then fix it at the last working version and then continue to increment the other until you get to the script's release date or all dependencies fail.
4. Record the dates that your software gets deprecation warnings and fails.

Another method:

Track a code bases through git commits and somehow measure the frequency and time of depredations and removals.

We will have to find a reliable way to get old dependencies installed. This is often quite a painful process to simply get things installed as they were from some point in the past.

Another thought:

We could check how many tests of a prior version raise errors or deprecation warnings.

Description

Prior Art

• 360 walking treadmill: https://www.actvr.com/

Proposed Methods

Required Resources

Description

• Add 3D simulations of the bicycles using ThreeJS
• Include an interface to specify the rider's parameters independently of the bicycle parameters
• Add the human controller from https://github.com/moorepants/HumanControl along with several typical bicycle manuevers for simulation of stable and unstable bicycles

Prior Art

• JBike6
• http://bicycleparameters.herokuapp.com/

Proposed Methods

Required Resources

Description

The idea is to select an existing bicycle that has handling qualities that you like and calculate it's eigenvalues/eigenvectors (i.e. simplest description of its dynamics). Then partially design a new different bicycle, for example maybe a small wheel short wheel base bicycle. Let some set of physical parameters be free parameters in the new design. The are the parameters the design would manipulate to get a desirable design for the small wheel bike. Now use optimization to find optimal parameters of the new bicycle design that minimize the difference in the eigenvalues/eigenvectors wrt to the existing bicycle design. This will adjust the physical parameters of the new bicycle design to try to get matching dynamics to the existing design even though you've fixed some aspects of the new design to be different than the existing design.

Prior Art

This paper gave me the idea:

M. Paudel and F. F. Yap, “Development of an improved design methodology and front steering design guideline for small-wheel bicycles for better stability and performance,” Proceedings of the Institution of Mechanical Engineers, Part P: Journal of Sports Engineering and Technology, vol. 234, no. 3, pp. 227–244, Sep. 2020, doi: 10.1177/1754337120919608.

They basically do this process but in a very manual way.

Proposed Methods

I think you can use the procedures I use in this paper to solve this problem:

https://bmd2019.figshare.com/articles/Expanded_Optimization_for_Discovering_Optimal_Lateral_Handling_Bicycles/9942938

But you'll have to be clever in designing the objective function to compare the eigenvalues/eigenvectors.

Required Resources

Description

Prior Art

Eldery using a cycling simulator: https://www.theguardian.com/world/2020/oct/02/the-global-cycling-competition-helping-care-home-residents-see-the-world

Proposed Methods

Required Resources

Description

Prior Art

Proposed Methods

1. Create (or find) a torso + arms model with muscles and use existing bicycle data to back out what muscles fire in the balance of the bicycle (could use MOCO, maybe inverse dynamics [but probably don't have enough force measurements], or the other builtin tool in opensim C...)
2. Put EMG on a bicyclist and have them balance. Plot the muscle activation and determine primary players.

Required Resources

Description

Marco and Joris presented what they've learned about causes of injuries and death related to bicycle incidents. One clear thing is that the data is very spotty and incomplete. It is difficult to say definitively what the causes are what causes are the one that should receive most focus. We need to address the data gaps and make the record more complete.

• what if hospitals had an improve survey/questionnaire to fill out when someone comes in from a bicycle crash. should there be more standardization to the questionaires?
• the non-serous injury data is absent, do we need this information, if so can we collect it somehow?
• are their ways to take the incomplete data and answer the questions we want with statistical confidence?
• can we merge all existing data in better database, is there a database?
• how do we merge data from different countries?

Prior Art

https://bikemaps.org/ is a crowdsource collection of data.

Proposed Methods

Required Resources

Description

Make a robot bike that can perform trials style maneuvers.

Prior Art

Proposed Methods

Required Resources

Description

Bicycle theft is a major problem in many locals around the world. We should figure out how we can help reduce bicycle theft. I don't have any grand ideas that people haven't likely thought of before but we should start with some serious research into what, why, how bicycle theft happens.

https://www.dutchnews.nl/news/2020/08/bikes-worth-e600m-were-stolen-last-year-as-organised-gangs-move-in/

https://www.amweb.nl/schade/nieuws/2020/09/anwb-haalt-inmiddels-6-op-10-gestolen-fietsen-terug-plus-3-miljoen-aan-heroine-101125146?io_source=t.co

https://www.dutchnews.nl/news/2021/02/e-bikes-are-popular-with-professional-thieves-thefts-rise-37-last-year/

Prior Art

Proposed Methods

Required Resources

Description

Recreational skiers and snowboarders are regularly injured on jumps. Ski resorts construct these jumps with little to no design. The jumps can be designed to be safer. For example, jumps designed with a constant equivalent fall height that is small are likely safer than those with large equivalent fall heights.

Prior Art

Here are some papers to start with:

[1] M. Hubbard, “Safer Ski Jump Landing Surface Design Limits Normal Impact Velocity,” Journal of ASTM International, vol. 6, no. 1, p. 10, 2009, doi: 10.1520/STP47480S.
[2] J. A. McNeil and J. B. McNeil, “Dynamical analysis of winter terrain park jumps,” Sports Engineering, vol. 11, no. 3, pp. 159–164, Jun. 2009, doi: 10.1007/s12283-009-0013-8.
[3] A. D. Swedberg, “Safer ski jumps: Design of landing surfaces and clothoidal in-run transitions,” Master of Science in Applied Mathematics, Naval Postgraduate School, Monterey, California, 2010.
[4] J. A. McNeil, M. Hubbard, and A. D. Swedberg, “Designing tomorrow’s snow park jump,” Sports Engineering, vol. 15, no. 1, pp. 1–20, Mar. 2012, doi: 10.1007/s12283-012-0083-x.
[5] M. Hubbard and A. D. Swedberg, “Design of Terrain Park Jump Landing Surfaces for Constant Equivalent Fall Height Is Robust to ‘Uncontrollable’ Factors,” in Skiing Trauma and Safety: 19th Volume, R. J. Johnson, J. E. Shealy, R. M. Greenwald, and I. S. Scher, Eds. 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959: ASTM International, 2012, pp. 75–94.
[6] A. D. Swedberg and M. Hubbard, “Modeling Terrain Park Jumps: Linear Tabletop Geometry May Not Limit Equivalent Fall Height,” in Skiing Trauma and Safety: 19th Volume, R. J. Johnson, J. E. Shealy, R. M. Greenwald, and I. S. Scher, Eds. 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959: ASTM International, 2012, pp. 120–135.
[7] M. Hubbard, J. A. McNeil, N. Petrone, and M. Cognolato, “Impact Performance of Standard Tabletop and Constant Equivalent Fall Height Snow Park Jumps,” in Skiing Trauma and Safety: 20th Volume, R. J. Johnson, J. E. Shealy, and R. M. Greenwald, Eds. 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959: ASTM International, 2015, pp. 51–71.
[8] D. Levy, M. Hubbard, J. A. McNeil, and A. Swedberg, “A design rationale for safer terrain park jumps that limit equivalent fall height,” Sports Engineering, vol. 18, no. 4, pp. 227–239, Dec. 2015, doi: 10.1007/s12283-015-0182-6.
[9] N. Petrone, M. Cognolato, J. A. McNeil, and M. Hubbard, “Designing, building, measuring, and testing a constant equivalent fall height terrain park jump,” Sports Engineering, vol. 20, no. 4, pp. 283–292, Dec. 2017, doi: 10.1007/s12283-017-0253-y.
[10 ]P. Piprek and F. Holzapfel, “Robust Trajectory Optimization of a Ski Jumper for Uncertainty Influence and Safety Quantification,” Proceedings, vol. 2, no. 6, p. 320, Feb. 2018, doi: 10.3390/proceedings2060320.
[11] O. Audet et al., “What are the risk factors for injuries and injury prevention strategies for skiers and snowboarders in terrain parks and half-pipes? A systematic review,” Br J Sports Med, p. bjsports-2018-099166, Aug. 2018, doi: 10.1136/bjsports-2018-099166.

Proposed Methods

Here are some starting points for projects:

• Simple, accurate, and inexpensive measurement of the jump shape. Start with the Sno Park Profiler by McNeil and develop improvements or a new and better system.
• Jumps are constructed with Sno moving machines (Sno-cats). Design a tool that allows the drivers of these vehicles to construct a specified jump shape with high accuracy. For example, if the blade of the machine is dragged over a surface, record the jumps dimensions and compare it to a desired shape.
• Prove that equivalent fall height is a measure of injury severity (if true). A rotating and flipping person that impacts the ground can experience high impact forces due to the impact velocity normal to the surface and due to impact tangential if friction is high. The goal is to determine a simple (if possible) measure of potential danger.
• Identify the best and simplest estimate of injury severity based on physical measurements of the jump.
• Develop http://skijumpdesign.info into a optimally useful design tool for laypeople.

Required Resources

Description

Prior Art

Snuffelfiets is some kind of Dutch data collection program for bicyclists: https://snuffelfiets.nl/ (focus on GPS and sniffing the air for pollution)

Snuffelfiets is mentioned here: https://dutchcycling.nl/en/bestpractices

The Dutch have an open fiets data platform:

• https://dexter.ndwcloud.nu/opendata/bicycle
• See the article too: https://dutchcycling.nl/en/news/blog/bikes-and-its-train-your-data-reflex

Proposed Methods

Required Resources

Description

Many early prominent mutlibody dynamics codes use computer algebra systems (CAS) to symbolically derive long form analytic expressions for the equations of motion of the system. Most modern multibody codes are of a purely numeric approach, where the equations of motion are formed and evaluated in the same process. The hypothesis is that code generated from the symbolic forms can be significantly faster than a numeric based system.

Prior Art

• https://en.wikipedia.org/wiki/Computer_algebra_system
• Mailing list and Github discussions in the SymPy and PyDy projects will be useful

Symbolic mutlibody dynamics codes:

• SD Fast
• Autolev
• Auto/Bike/Truck/CarSim (derived from Michael Sayer's work at University of Michigan), https://www.carsim.com/, http://www.umtri.umich.edu/our-results/publications/symbolic-computer-language-multibody-systems
• MotionGenesis
• SymPy mechanics
• add others to this list, there are many more

Papers that may be useful to this project in our Zotero group: https://www.zotero.org/groups/966974/mechmotum/collections/GB9UR7YK

• sympy2c: from symbolic expressions to fast C/C++ functions and ODE solvers in Python https://arxiv.org/abs/2203.11945

Proposed Methods

• Improve the Autolev to SymPy parser and get it working smoothly: https://docs.sympy.org/latest/modules/physics/mechanics/autolev_parser.html
• Take some of the more complex Autolev and SymPy mechanics benchmark problems (human arm model, bicycle, etc) and derive the EoMs to be used for evaluation efficiency tests.
• Research the necessary optimal symbolic forms of the EoMs, any per-compilation optimizations for code generation, and finally compilation for optimal run times.

Required Resources

• Decent laptop/desktop computer. Linux or Mac will be best for all the software needs.

Description

Cars contribute significantly to air quality and other environmental degradation due to the tires wearing and producing tire particles. It would be nice to know what bicycles contribute to this issue also.

Prior Art

Proposed Methods

Required Resources

Description

Here's a recent example:

https://www.chron.com/news/houston-texas/transportation/article/liberty-county-bike-crash-kent-wosepka-16582033.php

It would be nice to quantify this across the country in some way that tests the hypothesis that the drivers are systematically given no consequences relative to other similar offenses. Sad thing is that this pretty much goes for all car crashes. The general populace considers death and injury due to car crash as a necessary evil instead of a preventable evil.

Prior Art

Proposed Methods

Required Resources

Description

Or real time feedback to help you minimize your frontal area while riding.

Prior Art

• This company may have a live feedback of drag coefficient: https://twitter.com/BodyRocketCC

Proposed Methods

Required Resources

Description

The UCI banned the "super tuck" position claiming that it was unsafe:

https://www.velonews.com/news/road/egan-bernal-michal-kwiatkowski-question-uci-priorities-in-wake-of-super-tuck-ban/

But it is quite unclear what criteria they use to determine the safety of such a position. It would be interesting to develop some bicycle models with different rider (aero) positions and rider-bicycle connections and determine if there is any effect on the open loop stability. Extending to closed loop could be interesting too, but we need a decent control model for these odd positions, which is likely different than normal positions.

Prior Art

Proposed Methods

Required Resources

Description

Accurate time series of steer torque and a variety of kinematics of bicyclists performing closed loop maneuvers and perturbation recovery are available here:

https://figshare.com/projects/Human_Control_Identification_During_Bicycling/22067

We showed that the linear Whipple-Carvallo model was not an ideal predictor of the plant model in [Moore2012] and [Moore2013]. In these papers we identified 4th order plant models that provided better explanatory power, but the best plant model was a black box. It is likely that a simple extension or some simple extensions to the Whipple-Carvallo model could provide first principles explanations of the definciencies in the Whipple-Carvallo model. Ideally this new model would stay 4th order, because the black box 4th order model is predictive. Some things that could be investigated:

• simple tire scrub torque models

• stiffness, damping, and inertial effects of the arms

• stiffness, damping, and inertial effects of the neck/head

• friction in rotating joints

• [Moore2012] http://moorepants.github.io/dissertation/systemidentification.html#bicycle-model-validity

• [Moore2013] J. K. Moore and M. Hubbard, “Identification of open loop dynamics of a manually controlled bicycle-rider system,” presented at the Bicycle and Motorcycle Dynamics, Narashino, Chiba, Japan, Nov. 2013.

• [vanderKooij2005] H. van der Kooij, E. van Asseldonk, and F. C. T. van der Helm, “Comparison of different methods to identify and quantify balance control,” Journal of Neuroscience Methods, vol. 145, no. 1–2, pp. 175–203, Jun. 2005, doi: 10.1016/j.jneumeth.2005.01.003.

Prior Art

• SymPy/PyDy Whipple-Carvallo model pydy/pydy#122

Proposed Methods

Required Resources

Description

Can we create a website that curates time series and associated meta data from single track vehicle motion?

Types of data:

• IMU time series
• GPS time series
• Temperature
• Barometer
• Video (on bicyclists, bike, other)
• Sound
• Lidar
• Eye tracking

Prior Art

Proposed Methods

Required Resources

Description

Drivers often characterize bicyclists as habitual traffic rule breakers and use this to justify giving them lower priority and status on the roads. It would be nice to quantitatively know how often bicyclists break rules, what kinds of rules they break, how unsafe do breaking these rules make things for them and others, etc. But this also needs to be done for car drivers so that a comparative analysis can be done. My hypothesis would be that bicyclists and car drivers break rules a similar amount but that bicyclists break rules to ensure their personal safety in many of the cases whereas car drivers break rules more often for personal convenience.

Prior Art

https://www.forbes.com/sites/carltonreid/2019/05/10/cyclists-break-far-fewer-road-rules-than-motorists-finds-new-video-study/?sh=36a0b3834bfa

Proposed Methods

Required Resources

Description

It would be quite nice to have a website that gives overviews of the different open source multi-body dynamics software for comparison. It would ideally have low bias and if comparisons could be made of benchmark problems that would be quite nice. The benchmark problems should let people compare the code they have to write, how complex installation is, computational efficiency, etc.

Prior Art

• simbody
• dart
• chronos
• opendynamicsengine
• bullet
• pydy
• https://github.com/DocBO/mubosym
• etc

Attempt at a benchmarking website: https://www.iftomm-multibody.org/benchmark

Proposed Methods

Required Resources

Description

These are some notes from speaking with USA Skateboard Education Association.

Joe Eberling and Chris Hargrave

The Ollie

• discover how to teach an ollie faster and safer
• minimize injury
• analyze people that can't ollie to find what is missing
• I should watch coaching tutorials (most focus on one thing instead of all things).
• Ollie is good fundamental trick to study.

Speed generation in pumping

• How to pump a transition

• Tick tacking

• How to pump a bowl

• Example of extraordinary pump (Sandro Dias) bailed: one pump to get a 540.

• Progressive transition so that it goes from less steep to steep.

• How pressure distributes across the board due to foot position.

• Most common thing to cause injury: twig stopping wheel

• How many different foot positions for different tricks

Drop in

• How do we get people into doing that.
• Environment is wrong for learning.
• Aesthetics effect the emotions and fear.
• Design drop in ramp: progressive

Shape and intensity of features

• Optimal tranny shapes
• How many skatepark designers understand how to design the shape.

Grindline: builds skateparks

Rating system for skateparks

Framework for simple ranges of motion that are common throughout skateboarding. What are the extreme ranges in motion and what is the neutral "range". Chris focuses on four ranges: up/down, rotation, foot to foot, toe to heel.

Current: Joe is trying to learn backside 50-50s. Moving different steps: head, shoulders, carving into backside by weighting.

Board that shows pressure

Lateral learning: look to opposite wall.

Chris does a weekly call for snowboard instructors.

Prior Art

Kirsten and Michael Meyers

Proposed Methods

Required Resources

Description

We've now seen massive destruction of bicycles in China and from JUMP when they deem the products not useful. This seems to be a massive waste of resources and is a anti-pattern to believing that bike share is some kind of contribution to sustainability.

Prior Art

Proposed Methods

Required Resources

Description

Write up here: https://mechmotum.github.io/jobs/msc/determining-dynamics-perception-thresholds-of-bicycles.html

Prior Art

This theory is likely applicable: https://en.wikipedia.org/wiki/Weber%E2%80%93Fechner_law

Proposed Methods

Required Resources

New York City installed a large number of parking protected bicycle lanes over the last decade and many cities are now copying this design. But there is an alternative for a road with shared auto/bike/auto parking: the buffered bicycle lane. The parked cars stay at the curb, then a wide bicycle lane, and then a wide painted buffer. The buffer, bike lane, and auto parking could even be slightly raised from the auto travel lane. I believe that having the parked cars between the bicyclists and the autodrivers creates a visbility problem. Auto drivers have a appearing/disappearing bicyclists in the peripheral. I'd like to prove/disprove whether there is a significant safety difference between the two designs.

Description

We have at least two sets of data that could be useful for building a markerless motion capture algorithm for bicycling.

This data from bicycling on a treadmill using a traditional motion capture system:

https://figshare.com/account/projects/1860/articles/1082512

We also have the matching videos for this.

We also have videos and IMU type data for this: https://figshare.com/account/home#/projects/22067

There is video from a rear frame fixed camera and IMU type data from this paper too: http://moorepants.github.io/dissertation/delftbicycle.html

There may be more.

The idea would to use modern machine learning techniques to build an open source tool that can extract bicycle kinematics from videos.

Prior Art

This software is getting recent press: http://www.mousemotorlab.org/deeplabcut but was based on https://github.com/eldar/pose-tensorflow, the later which is specifically for humans.

Extracting gait data from cerebal palsy gait: http://gaitlab.stanford.edu/

Heike mentioned this paper about elderly falls as a helpful reference: https://pubmed.ncbi.nlm.nih.gov/23083889/

Openpose is used to extract kinematics of people.

Frame by frame analysis of horse racing falls: https://doi.org/10.1007/s12283-020-00323-0 (manual way of doing something similar to what we could do here).

2021-02: University of Gent says it is going to extract data from old vidoes https://www.vrt.be/vrtnws/nl/2021/02/03/ugent-onderzoek-wielrennen-valpartijen/, this guys is quoted in the article: https://ai.ugent.be/people/StevenVerstockt.en.html

https://twitter.com/JonMatthis/status/1351531974364688385, https://github.com/jonmatthis/freemocap

https://github.com/geaxgx/depthai_blazepose

https://anipose.readthedocs.io/en/latest/

Proposed Methods

Required Resources

Description

Add features to opty (https://github.com/csu-hmc/opty) that translate a generic class of cost functions into the form needed for the non-linear programming problem with a direct collocation transcription.

Prior Art

https://github.com/csu-hmc/opty

Proposed Methods

Required Resources

Description

Is it a bicycle or a motorcycle? Does it belong on the bike path or in the road? What is the answer to this?

You can force vehicles to behave a certain way (GPS speed regulators) or you can group them somehow.

Prior Art

Proposed Methods

Required Resources

Description

We need a system we can mount to most any bicycle that can applied lateral perturbations.

Requirements:

• portable (movable to different bikes, ideally attached in locations we don't necessarily always have other measurement hardware)
• should be able to apply up to 200 N impulses that are similar to the push/pull methods we've done before
• shouldn't add too much more mass to the bike
• we need to know it's inertial properties for easy addition to models
• should be able to do psuedo white noise, sum of sinsoids, steps, ramps, impulses, sines, and ideally follow arbitrary signal curves
• bandwidth? what should its transfer function look like? it would be nice to have a nice flat amplitude up to like 10 rad/s

Prior Art

Proposed Methods

Required Resources

Description

Saw this from TUD: https://doi.org/10.1038/s41467-020-18353-4. Very interesting driver model. Could be expanded to bicycling likely.

Prior Art

Proposed Methods

Required Resources

Description

Something along these lines could be done: https://www.reddit.com/r/dataisbeautiful/comments/j214ja/retinal_optic_flow_during_natural_locomotion_oc/

Prior Art

https://journals.sagepub.com/doi/pdf/10.1177/1477153514522473

Retinal optic flow during natural locomotion https://doi.org/10.1371/journal.pcbi.1009575 this is an update to a paper from 3 years ago. watch the video too: https://www.reddit.com/r/dataisbeautiful/comments/tb0fx1/gaze_and_foot_placement_when_walking_over_rocky/. We've got to do something like this on a bicycle in traffic.

Proposed Methods

Required Resources

Description

There is the long standing debate of whether bicycle helmets should be mandated by governments, similar how automobile seat belts are. If we want to make the act of bicycling safe, I see the dichotomy being:

1. in the event a bicyclist collides with something, ensure their injuries are minimized

vs

2. minimizing the chances that the bicyclist will be in a collision

Many societies have seemingly accepted that some # of collisions will occur in transportation. Lot's of money is spent in research, development, policy, etc. to make colliding safer for the collider. There seems to be considerably less attention in transportation to reducing the chance of collisions. The words "safe", "safety", "make safer", etc. are all used to describe 1) and 2) but there is a difference and using the word safe for both is a misnomer of sorts. 1) and 2) have to be dealt with, but I think the vast majority of resources should be put into 2) instead of 1).

So, can we contribute to this conversation from the bicycle lab? The hypothesis is that it would be best for societies to focus on 2) instead of 1) to improve safety the most. For the case of bicycles, this means more specifically to reduce focus on bicycle helmets. Can we offer anything to help prove this hypothesis?

Prior Art

Proposed Methods

Required Resources

Description

One longstanding issue is that there is no open source implementation of bicycle/motorcycle models that is accessible and usable. It seems that virtually all prior computational models are tied to proprietary things. The models are typically difficult to get a hold of an use.

Prior Art

Proposed Methods

Required Resources

Description

Given different vehicles: cargo bikes, bicycles, walking, vans, trucks, etc. figure out how a transport company can optimally use the different vehicle types in a package distribution scenario.

Prior Art

• https://pedalme.co.uk/why-cargo-bikes/
• 2021 Cargo bike hackathon hosted by Groningen: https://cargobikathon.nl/, this is related to: https://www.fietsdiensten.nl/

Proposed Methods

Required Resources

Description

Each person decides to remove their foot from the pedal and try to place it on the ground when they are going to fall (going too slow or a more distinct fall). Measuring when and why a person removes their foot from the pedal is a good indicator of the person's intuitive understanding that they need help with balance. Would be nice to quantify and understand this.

Prior Art

Proposed Methods

Required Resources

Description

It would be cool to have a collaboratively developed online book/website that is similar in nature to the Bicycling Science book(s). It would be a definitely resource for research backed scientific and engineering information on bicycles. Much like Sheldon Brown's website is essential for the bicycle trades-person this would be essential for the bicycle engineers, designers, and scientists.

Prior Art

• D. G. Wilson, Bicycling Science, 3rd ed. MIT Press, 2004.
• Cheung, Stephen S., Cycling Science, Human Kinetics, 2017.
• Glaskin, Max, Cycling Science: How Rider and Machine Work Together, University of Chicago Press, 2012

Proposed Methods

Required Resources

Description

Prior Art

Proposed Methods

Required Resources

Description

Prior Art

• https://youtu.be/EtI1wubS0b8

Proposed Methods

Required Resources