After a long hiatus from the hobby, I’m finally back with a new set of pictures to share! Back in March 2023, I was looking for a new hobby which wasn’t as physically demanding as climbing or biking. Around the same time, I also realized that I had been to so many cool places on my trips but I didn’t have pictures to show for most of them. Putting these two together, I decided to get back into photography after more than seven years since I last held a camera. To test the waters without spending too much, I bought a used Canon 60D with a 10-22mm ultrawide lens to complement my 18-55mm kit lens, 55-250mm zoom lens, and 50mm prime lens. Although far past its prime in 2010, the Canon 60D still holds up pretty well today and the pictures from this venerable DSLR are instantly distinguishable from even the best of modern smartphones. Nothing beats larger image sensors! I later switched to a Canon R10, but the majority of the pictures in this edition come from the Canon 60D.
I will never tire of seeing the granite monoliths in Yosemite. I visited the National Park for the second time on a solo trip in June 2023 and I was moved by the beauty of the park in a way I wasn’t the first time I visited. More has been written about Yosemite than I can do justice to in this blurb, so I will let the pictures do the talking here.
Acadia National Park was the second National Park I visited this year and I think this picture is my favorite from the whole trip! I love the house at the bottom corner of the image, graciously providing a subject amongst the foliage.
There’s not a whole lot going on in this picture compared to the other ones, but I like the way the ground, sea, and sky line up. I shot this with my wide angle lens to exaggerate the leading lines in the image.
I shot this at a wooden toys shop in the town of Bar Harbor. I like the way the birdhouses are almost falling on top of each other in this image. There’s definitely a story here the more you look at it.
That’s all for this volume! I plan to start posting more pictures of my work more regularly :-)
This blog post is a bit different. This one is a story about injustice which I have been dealing with for over two years.
Both Hertz and First Financial Asset Management have been harassing me for damages of a vehicle which I was not liable for. The sequence of the events concerning this case is as follows:
At 956am on March 7 2021, I rented a Chevrolet Traverse (license plate NFRT51) from the Hertz rental at 2951 Market St, Philadelphia, PA 19104. When I went to pick up the car at the rental agency, the employee present at the counter made me sign a document that the car had no damage before handing me the keys. Only when I went downstairs to pick up my vehicle, I noticed that the car had what looked like paint scratches on its front bumper. I was under the impression that this damage was already accounted for as Hertz made me fill out the paperwork before taking the vehicle. Nevertheless, I took photos of all sides of the car before taking the car out of the parking lot. I noted the odometer reading at 17708 miles when taking the vehicle.
My pre-rental inspection pictures. The existing damage to the front bumper cover is clearly visible.
I returned the car at 959am on March 8 2021. When I returned the keys, the Hertz representative at the counter did not come downstairs to inspect the vehicle in my presence. The odometer reading when I returned the vehicle was 17942 miles.
On 15 April 2021, I received an email from someone named Chad Miner at Hertz who claims damages of USD 1129.83 for damage concerning the front bumper. The email contained a document where the damage was reported on 18 March 2021 with the car’s odometer at 18290 miles. The report also contained a pre-rental inspection report dated 28 October 2020 with the car’s odometer at 11821 miles. In this pre-rental inspection report, the car’s front bumper is unblemished.
Note how Hertz’s ‘evidence’ for showing the car was in good condition was taken months before we rented it.
I contested the claims in the email with my own pre-rental inspection photos as evidence. Mr. Miner does not address these photos when I send it to him. He provides me with a customer care number (833-932-2073) which I call several times with no response. Mr. Miner then called me on my phone number on 19 April 2021. On this call, I refused to pay for the damages. Mr. Miner said my photos are not acceptable as evidence.
Over the next several months, I received several calls from First Financial Asset Management (FFAM) asking me to pay for the damages caused to the vehicle. I explained the situation and refused to pay every time they called. The frequency of these calls varied, but it was sometimes as often as once a week. Later on, the calls tapered off to once a month or less.
On November 1 2021 I received a letter from FFAM with an offer to resolve the indebtedness with a one time payment of USD 903.86 if I pay by 29 November 2021.
To set the record straight, I sent an email to FFAM on 11 January 2022 explaining the situation once again. This email goes unacknowledged by FFAM.
I continue to get calls from FFAM to make a payment at irregular intervals throughout 2022. I refused to pay on each of these calls.
On 19 May 2023 at 803am I received a call (phone number 833-784-0956) from Randy Jones at FFAM threatening me with legal action if I didn’t pay on that very phone call. I clarified what legal action meant in this context multiple times and Mr. Jones said it could mean a lien though he does not mention what property the lien would be against. I restate my case but Mr. Jones disregards my statements and repeatedly says I would need to pay or face legal action. Mr. Jones then offers me the option of a one time payment of USD 905 or USD 1129 paid in two installments. Out of panic and not willing to deal with a court case when I am completely unprepared, I chose the first option to make a single payment of USD 905 with my credit card. Mr. Jones initiates the payment after taking my Chase credit card details and billing address over the phone. He then mentions that the payment is flagged from my bank as fraud and tells me to resolve the matter with Chase while he’s on the line. Before I can do so, he reports that the payment has gone through and the case is now closed.
On 22 May 2023, I file a chargeback with Chase after realizing the debt collector has broken the rules cited at the end of this article.
On 12 June 2023 at 959am I received another call (phone number 833-784-0956) from Randy Jones at FFAM. Mr. Jones brings up that he has received notice about the chargeback from Chase and that FFAM will be proceeding with a) legal action, b) garnishment of wages, or c) cancellation of driver’s license (?). Mr. Jones states that this call is to warn me before I get pulled over in the future. Again, as I realized after 19 May call, these threats are empty and cannot legally be made by the debt collector without a case judgement.
My complaints with Hertz and FFAM in this whole situation are as follows:
Hertz did not correctly carry out the pre-rental inspection when handing over the vehicle to me. An employee should have walked with me to the vehicle and carried out the inspection in my presence.
The “pre-rental” inspection report for the vehicle showing the vehicle in undamaged condition was over four months old when I rented the vehicle. The “post-rental” inspection was more than a week after I returned the vehicle. If the vehicle was indeed damaged on my trip, why did Hertz only notice ten days later after the vehicle was driven more than 350 miles since I returned it? Moreover, I find it hard to believe that a rental agency as established as Hertz would only have pictures of the vehicle four months prior to my rental with the car’s odometer 6000 miles less than when I rented it. Does Hertz have no pictures of the vehicle at all in the four months before I rented it to show it was in good condition?
Both Hertz and FFAM disregarded my photographic evidence that the vehicle was damaged before I rented it without any explanation of why they felt the evidence was invalid.
FFAM has harassed me several times on the phone for over two years by even calling me when I was at work. I find this unacceptable.
FFAM did not even acknowledge my email dated 11 Jan 2022 when I attempted to explain the situation again. The CFPB even states that the debt collector must pause collecting the debt until the debt collector has responded to my dispute. 
My call with Mr. Jones on 19 May was very unpleasant. I only paid the amount under the threat of legal action and possible garnishment of wages. I have since learned that it is in fact illegal for debt collectors to garnish wages  so the threat was in fact empty. This has caused me a great deal of distress and I am very disappointed in the way FFAM has forced my hand to pay for damages which I did not cause.
The Pebble Cobbler is an annual gravel bike race held in Bakersfield, California. This year it was held on a cold morning on February 11. Participating in this race is right up there with the most crazy, most risky, most insane, and most exciting things I’ve ever done.
A gravel bike race is normally held on relatively flat terrain with the most technical feature being loose gravel in the place of asphalt. But this race was anything but normal. This 85km course had as much as 1640m of elevation gain and several technical features which I would’ve felt nervous riding on a hardtail, let alone a gravel bike. There were sections where the climb became so steep that riding up was impossible and the only option was to push, if not carry, the bike. The descents could take you through beautiful green pastures one moment and through rollercoaster like berms the next. This ride was epic and I am so glad that I took the plunge to take part in it as my first major bike race ever.
What went well
My only goal for this race. Despite almost bonking midway, I’m proud of myself for making it through!
I am very happy with the choices I made with what I brought with me on this ride. I took my Giant Revolt 2 and swapped out the stock pedals for Race Face Chester MTB pedals with Adidas 5.10 Freerider MTB shoes to go along with it. I also invested in padded cycling shorts and a jersey for this race. Even the choice of bottle cages turned out to be crucial, for many cyclists lost their water bottles after flying over bumps. I am also glad that my tires survived all the abuse without a puncture. Every bit of cycling gear I got with me made the journey a little bit easier and it all helped getting me across the finish line.
Bakersfield is a fairly long drive from the Bay Area so I had to figure out my way there and a night’s accommodation for the race. I originally planned to pick up my race packet the day before, but heavy Friday afternoon traffic on the way to Bakersfield delayed my arrival by two hours. Fortunately I was able to pick up my race packet early morning on race day without a hitch.
What didn’t go well
My training for this race was badly hampered by all the rain in the Bay Area leading up to the race. Dry weekend days were nearly impossible to come by so I settled for doing 40-50km road rides when the rain wasn’t too heavy. I also did a loop of Alameden Quicksilver on my gravel bike to get a better idea of how to handle the off-road descents on drop bars. Although these rides made me more comfortable with the length of the Pebble Cobbler, I had very little practice in dealing with the amount of elevation I encountered on the race day.
I use energy bars filled with carbohydrates, protein, and fat for fueling myself for most of my longer bike rides. However, these do almost nothing for replenishing electrolytes and sodium levels. This nearly cost me the race had it not been for the packets of Lays chips at the second SAG (Support And Gear) station 50km in.
I am very particular about the way my bike feels and I had spent a great deal of time in trying out different saddle heights and handlebar settings to make it comfortable for long rides. Though I had set it up for comfort and efficiency, I found that my bike fit was still inappropriate for climbs as my back really hurt on the biggest climb of the race!
My bike is equipped with a stock 2x9 Shimano Sora transmission with a 32/48 crankset and 11x34 casette. This means my transmission’s lowest gear ratio is 0.94, which is probably just a little too tall for the climbs. I noticed that although I could get up climbs much faster than others with lower gearing, I was having to work a lot harder for it as well. In some of the steeper climbs, I had my tires spin out. This became more apparent towards the end of the race when I was totally spent from the big climb in the middle.
Even at 85km and 5300 feet of elevation gain, the Pebble Cobbler is the little brother of the Rock Cobbler, a truely daunting course with 122km of mileage and 8100ft of elevation gain. Both the Pebble Cobbler and the Rock Cobbler share the first 75km of madness, after which the Pebble Cobbler diverges to a bike trail back to the start whereas the Rock Cobbler has another 50km in store. This year, there were 750 cyclists taking part in the Rock Cobbler and 250 taking part in the shorter format. I was under the impression that the competition in the Pebble Cobbler was probably not quite as high as that found in the Rock Cobbler which would only draw the most hardcore of riders.
Sector 1 (Start to 25km)
The race started off in four waves ten minutes apart, with roughly 250 cyclists in each wave. The Rock Cobbler had three of the four starting waves while the Pebble Cobbler only had one. The start was jam-packed and it took a full minute before there was enough space that I could start pedalling. The race started off on a two-way paved bike trail going by the Kern river. This was my first time riding in a group and I was pleasantly surprised at how easy it was to cruise at 30+km/hr on the flat ground in this section. After a brief hill climb and descent, the off-road section started out with a steep climb! Several people opted to push their bikes on this section but I was feeling great so I decided to power through all the climbs.
The doubletrack descent which followed was smooth and easy - a good way to warm up before the gnarly sections which came later in the day. Before I knew it, I came across the first SAG stop and I stopped for a few minutes to replinish my food and water supplies. I also needed to adjust my seatpost’s bolt, or else I would have to contend with the seatpost sliding down into the frame on its own. The Cobbler is famous for mini challenges during the race and at the first SAG stop, the challenge was to jump over a fire pit for…posterity?!
Sector 2 (25km to 55km)
Riding off the high of this being my first race, I got to the 40km mark in 2hrs13mins and I was feeling pretty good about my pacing at that point. My map indicated that I would need to brace myself for two big hills at 42km and 56km in so I reduced my pace as we got off the paved section. Then, the big hill started with an annoying 5% gradient which never seemed to let up. My back started hurting only a kilometer into the climb and the pain progressively got worse the higher I got. The only reprieve I got was when the dirt turned into mud higher up the mountain. It was impossible to pedal anymore so I did what everyone else was doing - pushing my bike up through the mud. My flat Freeriders were of little use on this part and I could only make it a couple of steps before sliding back one. The whole climb took me almost an hour and my shoes and tires were completely caked in mud. A bunch of us tried cleaning out the tires and brakes with the popsicle stick we got in the race packet, but it was of little use when a minute of rolling down the mountain would soil the tires again. I took a deep breath, steadied myself on the bike and started the most dangerous descent of the race. There was almost no traction at all from my tires on the surface and the occasional ruts in the ground nearly threw me off my bike. I was fighting for control and it took every mountain biking instinct I had to hang on. The traction and braking power was so poor that I might as well have been skiing. Once the descent mellowed a bit, the next climb started almost immediately after.
This turned into a real hike-a-bike affair with the gradient touching as much as 20 degrees in some spots. I was completely gassed at this point and I was dreading any more elevation gain. Even sitting hurt despite my padded shorts and the cushioned saddle cover! The 55km mark came up in 4hrs and I rolled into the second SAG stop just as I felt my legs cramping up. To my surprise, the packets of Lays chips which I had turned a dismissive eye to on the first SAG stop saved my race. I swear I could taste every crystal of salt as I wolfed down two packets. Never have chips tasted so good as they did then. With supplies replenished and electrolytes rebalanced, I was ready to take on the final 30km of the race.
Sector 3 (55km to Finish!)
The final sector of the Pebbler started off with a simple creek crossing. My pain in the rear was still no better than it had been at the end of Sector 2 so I did most of the initial part of this sector standing on the pedals. I chose my battles with the uphill sections and pushed my bike up some parts which would have been faster to ride through. Fortunately, the paved section in between helped me find my second wind and I was back in the race with the momentum which had been drained out of me in the last 20km. I was back on the saddle again. Not for long though, as the course had another wall-like climb in store. This was the steepest yet, with an average gradient of 26.8%, shooting up as high as 37.5% in some places! Riding this out was completely out of the question, even a 1:2 gear ratio would’ve struggled up this hill. The best thing to do would have been to carry my bike over my shoulder, but I simply didn’t have the strength to do so. There was a lot of traffic up the hill so it became a dicey affair of pushing the bike while trying to stay out of everyone else’s way. The reward for getting to a top was a remarkable view of Bakersfield and some of the best doubletrack fun I’ve ever had. The descent concluded with a twisty ride through a canyon with rollercoaster like berms. On a hardtail it would have been enjoyable, on a gravel bike it was a white knuckle affair. I nearly slipped off a berm by going too slow on one section. Shortly after, I came across a group nursing a rider who had broken his ankle taking one of the berms. Seeing the writing on the wall, I took it slow and walked my bike until the surface had evened out enough to ride normally. A few climbs and descents later and I was back on the pavement with only 15km more go!
Just before the Rock/Pebble Cobbler split, I came across the water ball pit challenge. I wasn’t looking forward to wading through ankle deep water and getting my shoes all soggy. If I had been thinking straight, I would have thought to take my shoes off before getting in. Although every km beyond the 70km point was uncharted territory for me, I knew I had just enough in the tank to make it through the final 10km. The sun was finally breaking through the clouds, the path was paved, and the worst of the climbs was well behind me. I was feeling optimistic about finishing this and also proud about how far I had come in the race. After the course split, I found myself riding along the Kern river, just as I had a few hours ago at the start of the race. I was putting more into the pedals now and I was cruising at 27-30km/hr till the end. I found myself getting passed by some riders who had saving their final burst of energy for the end. I put my head down and watched the kilometers tick away on my watch. It wasn’t until I reached the car parking lot that it hit me that I was only seconds away from the finish line! But wait! The track had been rearranged while I was out at the race and the final sprint to the finish involved a few obstacles like riding through a sandpit, dodging barrels, and negotiating tree roots. I took my another right turn and I was right back where I was 6hrs18mins ago. I smashed the pedals one last time with my right fist pumping the air as I went pass the finish line, 149th . I could not believe what I had just experienced. I was hurting all over and I was only able to stand because of the adrenaline. My shoes were soggy and my clothes were all muddy, but I couldn’t care less in the moment.
It’s funny how something as natural as the progression of time still catches me by surprise. Nowadays, I frequently catch myself thinking that it was only last year I had freshly graduated from UPenn and left the student life behind me. It also takes me by surprise that it took me over this long before I started working a full-time job at a ‘real’ company. Living in the Bay Area and working full time has its benefits and downsides below:
Yes, the cost of living in the Bay Area is high. But tech jobs in the Bay Area pay really well, given my living expenses. More than anything, I feel like I’ve finally graduated from the student life when I don’t flinch at paying for a $15 entree. In the US, money gives you freedom and I feel like I’ve had more freedom living here than I’ve ever had before.
+ Learned to contribute meaningfully to a big company 🏢
I never worked at a ‘real’ company before I joined NVIDIA. Even coding collaboratively was new to me when I joined as all my projects before working at NVIDIA were solo projects under a mentor. As opposed to personal projects where I handled the design and implementation, I’ve noticed that it’s quite different to work on a smaller part of a big project where the scope of new designs is smaller. I’ve learned a lot about writing better C++ code and even picked up a little bit of CUDA for prototyping parts of our code on the GPU.
My first year of working at NVIDIA can be divided in two distinct phases. In the first phase part I worked from home all the time. And in the current phase, I’m in the office everyday. Working from home felt alienating in the beginning and I frequently questioned myself if this is what I’ve wanted since forever. With no company-wide events in 2021, it barely even felt like I was working for a major company. FH felt like going to a party where you can only watch others through the door’s keyhole.
Things have definitely changed for the better since the campus opened this year. I am much happier to go through the routine of getting ready and commuting compared to spending all day at home. The office is still mostly empty but I’ve made new lunch buddies and climbing partners along the way, which brings me to my next point about…
+ Climbing 🧗♂️
After spending a few hours in a climbing gym, it’s easy to see why indoor rock climbing is so popular in the By Area. The climbing routes (called problems) tickle the same parts of my brain as programming does. Problems are graded on a scale by difficulty. Some problems require special moves. And you’re constrained by the time you have to solve a problem by how much your fingers and forearms can take before you find yourself getting more familiar with the climbing pad than the top of the wall.
The graded progression in problem difficulty in climbing gives me a lot of motivation to keep at it. I love that it’s also a good workout while barely feeling like one. I was also pleasantly surprised to find that the indoor rock climbing community is very inclusive. I’ve found that discussing the beta is one of the best ways to strike up conversation and make friends with complete strangers.
+ More extroverted 🤗
I feel like 2020 really did a number on my social skills. When the world was slowly opening back up again, I felt like I was starting afresh. I was finding my way through learning how to interact with people outside of my COVID buddies group, and suffice to say, there were some cringey moments along the way. Anything more than a cursory ‘thank you’ at the grocery store felt foreign to me. With inexperience comes opportunity and I felt excited to speak to people in a way I never had before. Social events which I would’ve avoided in the past were now unmissable for me. Personally, I think this is the biggest change I’ve seen in myself since I came to the US in 2019. It definitely helps to have something interesting to talk about when meeting someone new, and I feel like accumulating several experiences and stories over the years has helped me a lot in this regard.
+ Settling 🏠
I didn’t realize it at the time, but my first year in the Bay Area was a process in getting more settled. I used to hate the term, because I used to conflate settling with slowing down. I found staying at home difficult in the beginning - for it was the last thing I wanted to do after a year of online masters and a WFH job. But settling in is more than just staying at home - I’ve come to realize that settling is feeling more secure in your lifestyle.
It’s about being more sure about where you want to be in this stage of life. Getting settled is a process which I never paid much attention to in Hyderabad, Singapore, or Philadelphia, because I always had it at the back of my mind that I would be moving eventually. It’s different now as the habits I form now are going to stick for much longer. To this end, I’ve been working on scheduling appointments, keeping track of chores, and maintaining a personal kanban board for things I want to accomplish in two week intervals. One of the most effective tricks that’s worked for me is to resolve the daily sources of friction I used to put up with at the moment I identify it.
Working from the office has helped bring some balance to life and has restored the home to a place meant for rest. I’ve made changes to make my room more livable by spending some time in reorganizing the furniture. These small changes don’t take much time to implement, but their effect is felt everyday.
- Maintenance 🔧
One thing I’ve recently learned from personal experience is that I could enjoy a ‘high-maintenance’ lifestyle on the cheap by trying to become more up to date on chores. It comes with trying to be more settled and aiming for a higher quality of life. Yet, living in the US means that chores are a time-consuming endeavor. I’ve tried to rotate through a number of maintenance related tasks throughout the week, but it feels never-ending! The worst of it is unexpected maintenance in the middle of the week throwing my schedule off. This happens more often than not with my transportation, e.g., bumper damage on my car or a puncture on one my bikes.
- COVID Rebound 🎢
It took me a year to realize it, but my rebound to life after COVID was not sustainable. I tried to maximize the time I was out of the house on weekends and weekdays after the office opened. It was fun initially, but the routine of spending 12 hours a day on the move caught up with me, which is why I then realized I need to improve the baseline quality of life by settling in a bit more.
There are some rare events in your life which you know will be the turning point for what comes ahead. ICRA 2022 was one of those for me.
ICRA, the International Conference for Robotics and Automation is the largest robotics conference in the world and witnesses thousands of attendees from dozens of countries. This year’s ICRA was particularly special as it was the first time ICRA was held in person since ICRA 2019 in Montreal. It was special to me too as it was held in Philadelphia right after my graduation ceremony from the University of Pennsylvania. The timing could not be better and I am delighted that NVIDIA fully sponsored me to attend the whole conference comprising of two days of workshops and three days of exhibits, technical talks, and poster presentations. While this isn’t my first international conference, it is my first academic conference.
There was little reason to doubt the scale of the conference when I saw the length of the line for registration at 830am on the opening day. It took me a good 5 minutes to walk to the end of the line, which wrapped around the ground floor of the enormous atrium of the Pennsylvania Convention Center. Standing in line gave me a good chance to survey the crowd, comprising of a mix of industry professionals and PhD students. I guess the professors and post-docs had the good sense to collect their badge from the convention center the previous day since I didn’t see any waiting in line. On reaching the front of the line, I collected my ICRA kit containing my conference badge, an ICRA branded bag, three adorable stickers for my badge to indicate whether I preferred to elbow bump, shake hands, or hug, and (what would become a theme for the conference) an ICRA T-shirt!
I attended the workshops on novel perception techniques and autonomous driving. I also got a sneak peek of the company exhibition booths on the second level of the convention center. Apart from being a great place to meet researchers, I learned ICRA also serves as a recruiting event for several startups and more established robotics companies such as Skydio, Dyson, Motional, and Tesla. Over the following days, I amassed enough company swag to cover my clothing needs for the entire the conference. Had I known better, I would have probably packed a little lighter.
The day ended with a huge networking dinner in the ballroom of the convention center. Over dinner I was pleasantly surprised to meet old friends who I didn’t even know were attending and even one of my TA’s at UPenn who had successfully raised millions of dollars in funding for his autonomous lawnmower startup, Novabot!
The autonomously lawn mowing Novabot, developed by my CIS581 TA Yulai Weng!
It had been such a long time since anyone had attended an event of this size and it was clear that everyone was relishing the opportunity to socialize and geek out. For someone who wouldn’t normally walk up and talk up to people unsolicitedly, I found myself really enjoying the process of introducing myself and learning about what other people were working on.
Networking dinner at the end of day 1! Where’s Waldo?
The next three days formed the main part of the conference with technical presentations, company exhibitions, and boozy recruitment parties. This year’s ICRA adopted a hybrid model for the paper presentations. Compared to the first ICRA in 1985 which had one track with each 25 minute presentations, ICRA in 2022 had 22 parallel tracks with 5 minute lightning talks in the morning and afternoon. I felt 5 minutes was barely enough time to grasp the keywords of the paper, especially if it was in a field which I hadn’t worked in before. Fortunately, the lightning talks were followed by in-depth poster presentations where you can walk up and ask the authors questions about their work. I found great value in the poster presentations and ended up learning a lot by talking to the authors.
As an aside about the hybrid model of presentations, some authors chose to present using a 5 minute video of them speaking over their slides and answering questions over Zoom. It was somewhat ironical that we had gathered from thousands of miles away only to watch a video together.
My school friend’s, Aditya Arun’s, poster presentation!
I also loved attending the company exhibition booths in between the paper presentations, especially those where engineers from the companies came to describe what they were working on. Based on the booths, my impression of the robotics industry is that it is dominated by a new wave of software companies either 1) leveraging existing hardware (particularly UAVs and service bots) for new applications or 2) providing tools for other robotics companies. Companies like Skydio, Shield AI, and OhmniLabs fall into the first category whereas companies like Foxglove and Tangram Vision fall in the second. What was less common was companies providing hardware (such as sensors and motors) to companies in the first category such as Ouster and DirectDrive motors. The exhibitions piqued my curiosity for how these startups discover a market niche and create products to fill said niches.
The second day of the conference ended on a high with the Skydio recruitment party at a bar in Philadelphia’s Chinatown called bar.ly. Over beer, I chatted with some European post-docs and PhD students who had a hand in designing the Mars helicopter, Ingenuity. Ingenuity has to contend with some of the most outlandish conditions any robot has experienced and has to sustain aerial flight in an atmosphere 1% of the thickness of Earth’s. Roughly speaking, that equates to flying a drone at an altitude of 100,000 feet above Earth’s surface (albeit in lower gravity). I also learned that robot localization is much more challenging on Mars, for the planet’s lack of a magnetic field makes magnetometers useless and for the time being at least, the planet is bereft of a GPS/GNSS network. Surprisingly, that makes visual inertial odometry (VIO) the most effective way to localize Martian robots. However, Mars’s surface is devoid of significant visual features which VIO algorithms heavily rely on. To this end, the VIO algorithm for Ingenuity was tested on deserts on Earth which lack significant visual features the way the Martian surface does. As Ingenuity is experimental hardware, it can use the consumer-grade Snapdragon 801 for its flight computer whereas other space robots use much slower, radiation hardened processors for better reliability.
ICRA also had some great keynote talks with Julie Shah opening Day 3 with a talk about what the future of robots and humans working together might look like. It never occurred to me that safety for the humans and task efficiency for the robots working with them could be conflicting goals. Research in this area deals with designing ways robots can better predict the actions of humans and how they can offer social cues when working alongside them. Another challenge is making complex machines like robots easy to train and maintain for operators.
Me at the morning keynote session!
In a more formal setting than the discussions I had about the same topic the previous night, the afternoon keynote talk was by Prof. Vandi Verma at NASA, who spoke about the challenges of space robotics and the development of Ingenuity. I learned that the only picture of the Perseverance rover and Ingenuity in the same frame was shot with painstaking composition and caution, since the robotic arm carrying the camera has to be aligned correctly with the rover’s other instruments. Who would’ve thought that something as simple as taking a picture involves hours of simulations back at NASA?
Ingenuity and Perseverance in the same frame! Hours of planning went into this shot.
I also watched the semifinals and finals of the F1Tenth racing competition in which teams race self-driving 1:10 scale cars on an indoor track. I have particularly fond (and painful) memories of hacking together our own autonomously driven racecar in my second semester at UPenn. It was super impressive to see how far the project had come along with all the top team’s approaches having robust controllers, overtaking maneuvers, and even blocking maneuvers to avoid overtakes. I daresay that it’s even better than watching the ‘real’ F1 races! UPenn’s very own Team ScatterBrain brought the trophy home in a head to head final race with Dzik from the Czech Republic.
Johannes Betz commenting on the F1/10 races!
The day ended with a banquet dinner in the terrace of the convention hall with speeches from Prof. Vijay Kumar at UPenn and Spot from Boston Dynamics raising a toast to all robots (and humans behind them)!
Spot raising a toast to all the robots out there!
The penultimate day of the conference had technical presentations on topics which I work closely on at work - autonomous vehicle navigation and motion planning. Some of the presentations which stood out to me were 1) StopNet for whole scene prediction modelling from Waymo, 2) trajectory generation to maximize information gain in the presence of occlusions and 3) PredictionNet from our very own prediction team at NVIDIA. It was interesting to compare the differences in StopNet and PredictionNet, for the former is a sophisticated prediction model which shows what is possible when you have significant computational resources (more than 1TFLOP per inference!) to throw at the problem whereas the latter is an efficient prediction model which is designed to run in a few milliseconds on the NVIDIA Drive AGX platform. I also came away with the impression that its genuinely difficult to produce good quality research for autonomous vehicles at the university level. Several papers only had results in simulation which is hardly sufficient for demonstrating the validity of an approach. An end-to-end model which goes directly from camera pixel values to steering wheel torques may look impressive when demoed in CARLA, but it seldom holds water when implemented in a real vehicle. For better or for worse, the autonomous vehicles testing pipeline requires trained test drivers, testing tracks, expensive sensors, and of course, cars, which are expensive to retrofit with sensors and specialized hardware. This is very capital intensive and out of the reach for all but a few universities. With some bias, I am of the opinion that for this reason, most of the useful progress in the field will be led by the industry. On that note, the conference session ended with a panel discussion on the future of the AV industry with research leads from Tesla, Motional, Zoox, NVIDIA, Toyota Research International, Waymo, and Qualcomm chiming in. And as on brand for this conference, the night ended in more revelry with Zoox footing the bill for drinks and finger food at the Graffiti Bar in Philadelphia’s Center City.
Robots (and exoskeletons)!
With the technical talks and exhibition booths wrapped up on Thursday, the final day of the conference felt subdued compared to the ones before it, though this is in no small part due to the action of the last handful of days catching up to me. I flitted between workshop rooms to learn more about the latest in neural geometry based motion planning (think NeRFs) and autonomous driving. Lunch was followed by one of my favorite events at ICRA in recent years - the robotics debates. I attended the debate on ‘The investment and involvement of industry giants in robotics research in academia is positive’. At the outset, I felt it would be hard to argue against this topic and voted ‘Yes’ in the pre-debate polls. I was proven wrong by Prof. Ankur Mehta who made an excellent case for how in some ways, the funding offered by huge companies can create perverse incentives for academic research labs to pursue avenues of research which are more oriented towards creating profit for big tech companies. The relation between industry funding and universities is not always as synergistic as it may seem. Personally, I still feel that the industry and universities have benefitted from each other’s goals, but I am more on the fence about if it’s always a good thing. Academia and the robotics industry will always be competing for the best minds to join as PhDs or postdocs in the former’s case and as skilled engineers in the latter’s. What kind of systemic change is required so both can benefit without being at odds with each other?
At this point, my brain had pretty much checked out of ICRA and I attended the final workshop on motion planning under uncertainty, but couldn’t do much more than passively listen.
Despite being physically and mentally tired on the last day, I left the conference feeling very inspired. It’s hard not to be when you get time to interact face-to-face with people who are so venerated in the robotics community. There is something special about the energy of having thousands of people with the same interests together in a single room and I now understand why these conferences are such a big deal. It was a humbling experience too - for it showed me that there was so much I didn’t know in the very field I spent two years getting a masters degree in. I am bullish for the future prospects of robotics as ICRA has showed me that academic and industrial interest in solving the biggest open problems of robotics is at an all time high. People are seriously thinking about the inevitable ways in which our society will change once robotics become much more commonplace. And finally, automation could unlock huge markets of opportunity which didn’t exist before.
On a more personal note, I feel like my trip to the East coast this time has done me a world of good. Last year, I had written about how much I was looking forward to moving out of Philadelphia, but this time, I realized how much I had missed the city. Now that it’s summer, Center City and its numerous parks are filled with people and there is much joy to be found on the streets. The city has a charm which hasn’t been lost on me at all. I long for the spontaneity of urban life when living in the dull suburban sprawl of the South Bay. I feel like the 3000 miles of distance has given me a new perspective to reflect on life since moving to California. I have realized that despite having great friends nearby and a job I look forward to everyday, it all feels strangely devoid of contentment - something I say knowing full well that it comes from a position of immense privilege. That said, the general feeling of discontent is something echoed by the few with whom I’ve dared to broach the topic with as well. This is where I feel the conference was could be a turning point. ICRA has shown me that I feel at my best when talking to people and thinking about robotics outside of work. This is something I which I want to channelize in my weekdays as well, for I feel it’s all to easy to get stuck in a loop of expending one’s mental faculties at work and not leaving room for creative pursuits. There is so, so much magic left in robotics that I simply can’t hope to cover all of it. In a way, it’s a beautiful thing to be in a field where everyone can pick their own set of skills to be unique.